ARTICLE

A cryptanalytic decipherment of the Indus script

yajnadevam


ARTICLE HISTORY
Compiled November 7, 2024

Abstract
Indus inscriptions hold the key to unlocking the history of pre-Iron Age India and
all Indo-European peoples but remain undeciphered for over a century. All prior
attempts have been partial, unsatisfactory and unfalsifiable. We decipher the In-
dus script by treating it as a large cryptogram as described by Claude Shannon.
We decipher every sign sequentially using regular expressions and set-intersection.
Indus script is discovered to be proto-abugida segmental with signs for consonants
and vowels. Indus inscriptions are in grammatically correct post-Vedic Sanskrit.
Variants of 76 allographs constitute most signs. Conjunct signs constitute the rest.
Our decipherment can read every inscription and we translate 500+ inscriptions in-
cluding the 50+ longest, 50+ shortest and 400+ medium-sized inscriptions including
100+ inscriptions with conjunct signs. We comfortably surpass Shannon’s criteria
for a credible cryptogram decipherment. Brahmi glyphs are discovered to be stan-
dardized Indus signs. We find significant continuation of Indus linguistic features
and cultural elements in post-bronze age India.

KEYWORDS
Indus Valley Civilization; Indus script; Epigraphy; Brahmi; Sanskrit

1 1. Introduction

2 Indus Valley civilization was the largest bronze age civilization, spanning over one mil-
3 lion square kilometers and having an estimated population of five million (Dixit, 2019).
4 This advanced civilization featured planned cities, drainage, international trade and
5 standardized weights among many impressive accomplishments. Indus civilization has
6 left behind strings of symbols on seals, tablets and a large sign (Dholavira signboard)
7 which is termed “the Indus script.” Although the mature Indus civilization is believed
8 to have started around 2600 BCE, we see stage-by-stage evolution from the earliest
9 sites to Bhirrana and Mehrgarh 7000 BCE. These older stages are given different names
10 such as pre-Harappan, early-Harappan and so on to distinguish them from the mature
11 Harappan stage.

12 1.1. Geographic and temporal spread

13 The earliest Indus seal is from Kunal, dated to 4000 BCE (ASI, 2004). The earliest
14 graffiti appears on potsherds in 4000 BCE Balakot (Ahmed, 2014; Lashari et al., 2020).
15 Seals with Indus script begin in the archaeological record mid-to-late 4th-millennium
16 BCE (BBC, 1999). Seals post-1900 BCE are rare in the archaeological record. Indus
17 sites in the Thar desert are abandoned but sites in the Himalayan foothills continue till
18 about Jognekhera 800 BCE (Haryana Directorate of Archaeology and Museums, 2023).
19 Indus symbols however continue to appear on Varanasi coins from 800 BCE to Maurya
20 era coins and on megalithic sites in South India to 50 CE (Banerjee and Rajan, 1960;
21 CoinIndia, 2010; Reddy and Sakunthalamma, 2023) and Janapada seals and coins upto
22 350CE. B. B. Lal noted that 89 percent of megalithic symbols go back to Indus symbols
23 (Lal, 1960). Findings at Keezhadi show usage of both Brahmi and Indus symbols with
24 primarily Indus graffiti from 600 BCE and transitioning to only Tamil-Brahmi around
25 100 BCE (Ramakrishna et al., 2018; Sivanantham and Seran, 2019). Sealings continued
26 to appear until the Gupta era using Brahmi script (ASI, 1906). Mixed Brahmi-Indus
27 scripts occur in OCP copper hoards (Munjal and Munjal, 2005), Iran (Lahafian, 2013),
28 Tamil Nadu (Subrahmanian, 2010), Sri Lanka (Raghupathy, 1987) and Vietnam (Lien,
29 2013), ranging over many hundreds of years.

30 1.2. The corpus

31 There are 4300+ inscriptions from archaeological digs (Wells and Fuls, 2023). More
32 are being unearthed as new sites are excavated. The stratigraphy, location and other
33 details of most later excavations are preserved but many older finds have missing data.
34 The median inscription length is about 5 signs. There are about 50 inscriptions that
35 are 10 signs or longer. Some inscriptions are a single sign or two signs. The seals
36 themselves are small and come in all manner of shapes: square, rectangular, circular
37 and triangular. The square seals are sized about 3×3 to 4×4 cm. Due to the scarce
38 space, we see the crowding of letters to the left side in early inscriptions. The script
39 has a large variety of signs that are hand carved and many signs show abstraction,
40 simplification and rotation, evidently to preserve scarce horizontal space. The oldest
41 layer of Harappa shows only about 71 signs, possibly indicating that some later signs
42 are variants (Konasukawa, 2020).
43 In addition to seals, we have tablets and tags, that were presumably used to attach
44 to goods for trade. These show heavy wear similar to old coins and are found close to
45 market areas, which may be taken as evidence of use as currency. Many seals indeed
46 resemble Gupta seals and early Janapada coins with an inscription on top, an animal
47 icon in the center and sometimes a staff or standard in front of the animal (Srivatsava,
48 2021).
49 The Dholavira signboard is meant to be seen from a large distance to convey informa-
50 tion to travelers and may be evidence that the Indus script is a form of writing. Writing
51 on post-fired pottery, personal items such as tools, weapons, bangles and jewelry may
52 indicate that IVC was a literate society. Indus characters have also been found carved
53 in the Kirthar mountains of Sindh (?). The broad geographic and temporal distribution
54 along with mixed Indus/Brahmi inscriptions should be sufficient evidence that Indus
55 script represents writing rather than proto-writing or pictorial identifiers.
56 Although the Indus inscriptions are short, at least one example of a long inscription in
57 an evolved form of the Indus script exists in the form of the Vikramkhol cave inscription.
58 This inscription dated 1500 BCE is considered mid-way between the Indus and Brahmi
59 scripts by archaeologist K. P. Jayaswal (Jayaswal, 1933).

2
 60 2. Methods

61 2.1. Script Analysis

62 The oldest evidence of the Indus script is graffiti on potsherds from 4000 BCE Balakot.
63 There are both abstracted and elaborate versions of signs in this phase. Whether the
64 script originally began as a pre-writing system similar to cuneiform cannot be deter-
65 mined using artifacts from this phase. The script seems to have evolved into a uniform
66 writing system over a millennium. By the mature phase ~2600 BCE, inscriptions are
67 fairly consistent over time and space.

68 2.1.1. Script direction

69 The preferred script direction seems to be right-to-left as evident from external and
70 internal evidence (Mahadevan, 1977). External evidence such as crowding on the left
71 side on inscribed copper objects, bas-relief tablets, seal impressions (right side on seals),
72 evidence of starting carving letters from the right side and so on was combined with
73 internal evidence such as identical inscriptions appearing as both one and two line
74 versions. This enabled the eventual determination of the direction of most inscriptions.
75 Approximately 83% are determined to be right-to-left and 6.5% are left-to-right. The
76 rest are either single-sign, undetermined or damaged. A mathematical model of the gini-
77 coefficient of the Indus corpus also finds that the Indus script is written right-to-left
78 regardless of the type of script (Ashraf and Sinha, 2018).

79 2.1.2. The signary

80 Mahadevan classifies the Indus script into 417 signs, ICIT classifies them into 700+ signs
81 (Wells and Fuls, 2023), treating minor variants as separate signs since a sizable number
82 (1000+) of signs is suitable for a logosyllabic analysis. On the other extreme, S. R.
83 Rao claimed that the script only had 20 or so signs (Rao, 1980). The obvious problem
84 with a low sign count is that the number of legal words formed by adjacent signs drops
85 precipitously. There will be many instances of repeated syllables such as kakajakaja,
86 making reading anything but the shortest inscriptions impossible. Rao was forced to
87 claim that the Indus Valley language was monosyllabic. Essentially he acknowledged
88 that he could not find even a small number of legal words formed by the interlocking
89 of his assigned values. Maximizing the signary size increases the possible matches of
90 legal words. For example, if we have a partially deciphered string πAT , we may choose
91 to assign C to π to read CAT . However, if there is another sequence ΠπT , we get
92 a higher readability by choosing A for Π to read ACT , even though the signs π and
93 Π look similar. Logosyllabic analysis also benefits from a large signary by claiming
94 related meanings for similar signs, such as Parpola’s star  vs planet  for variants of
95 the fish sign. This is a workaround to avoid nonsensical logosyllabic readings such as
96 ‘star-star-jar’, which may now be read slightly better as ‘star-Saturn-jar.’
97 However, a large signary is untenable on thorough analysis. Inscriptions up to Harap-
98 pan period 3A show only 71 signs (Konasukawa, 2020). It is improbable for a large
99 signary script to exist from 4000 BCE to 2600 BCE with only 71 signs attested. On
100 examination, we find that many allegedly different signs that are attested post the
101 Harappan period 3A are cursive forms    , simplified abstractions     ,
102 embellished   and stylistic variants    or strings of pre-existing signs . Sim-
103 plified forms of complex signs are normal in long-lived scripts when the medium is not
104 conducive to cursive development, for example as in Sumerian cuneiform. Cursive forms

3
105 naturally evolve on soft writing materials like paper or cloth using ink due to the effi-
106 ciency of not lifting the writing instrument off the writing surface. This efficiency does
107 not retrofit to carved inscriptions, so non-cursive forms continue to be used on carved
108 medium, even to the present day. We can trace the evolution of cursive forms in Indus
109 sign variants that gradually minimize the number of strokes needed to render the sign.
110 This may indicate that the Indus script was evolving to cursive form before and during
111 the mature Indus phase.

112 2.1.3. Script Type

113 Logographic scripts are characterized by low frequency and a large signary. Only a small
114 fraction of the total 417 Indus script signs (27.5%) are attested 10 or more times and
115 only a tiny fraction (6%) have an attestation count of 100+. A significant fraction of
116 the rare signs seem to be stylistic variants, accidentally mirrored signs, cursive forms or
117 word fragments. Gradual abstractions of pictorial signs such as     must be seen
118 as the development of cursive variants of a single sign. The rarity of cursive sequences
119 like    indicates that these are simply strings of their component signs rather than
120 completely unrelated signs. With these considerations, the sign count drops to under
121 a hundred and the script is unlikely to be logographic. This strengthens the case for a
122 syllabic or segmental script.
123 The pattern of repetitions of the same sign multiple times in the same inscription
124 also supports syllabic or segmental (abugida/alphabetic/abjad/phonetic) script. A non-
125 script should be expected to see repetition in the use of symbols similar to the birthday
126 collision effect. This number crosses 50% when the number of symbols in a single
127 inscription is 25 or more. No Indus inscription is this long, so we wouldn’t expect to
128 see any repeated symbols.

 
417!
(417−25)!
1− = 0.52
41725
129 Birthday collision, of course, applies to uniformly random events, while segmental
130 scripts are not uniformly random due to phonotactic rules unique to every language.
131 Therefore more frequently used symbols should see more collisions in smaller-length
132 inscriptions than less frequently used symbols. We see repetitions in the shortest of
133 inscriptions to medium to long inscriptions. About 17% of inscriptions have a repeated
134 sign if we include immediately repeated and bracketed signs. The jar and fish signs are
135 prolific repeaters and the spoked wheel sign occurs 4 times the Dholavira inscription.
136 This is additional support for a non-logographic script. This number is likely to be
137 much higher once allographs are discovered. By comparison in a logosyllabic script,
138 repetitions are extremely rare in short inscriptions. For example, of the 488 subway
139 station names in Beijing, only 3 (0.6%) have a repeated Mandarin sign and none with
140 3 or more occurrences (wikipedia, 2021).
141 Words that start with doubled consonants are rare in all languages and certainly
142 absent in Sanskrit and Dravidian. The occurrence of inscriptions starting with a doubled
143 or tripled sign in inscriptions like M-1794  and L-105  is evidence that the script
144 is not alphabetic, but either syllabic or segmental abugida.

4
145 2.2. Cryptanalytic Decipherments
146 Some scripts such as the Copiale cipher have been deciphered using cryptanalysis
147 (Knight, Megyesi, and Schaefer, 2011). The corpus of inscriptions in an unknown script
148 represents the ciphertext and the source language represents the plaintext. The key
149 is the assignment of the script signs to its values. The output of the cryptanalysis of
150 the script is its decipherment. This model has worked several times in history. Ephron
151 was able to re-decipher the Ventris-Chadwick decipherment of Linear-B using crypt-
152 analysis (Ephron, 1961). He only found one sign that had a different value, which was
153 later acknowledged as correct by the original decipherers, showing the superiority of
154 cryptanalytic decipherments.

155 2.3. Cryptograms

156 A syllabic or phonetic script can be modeled as a cipher and solved using proven math-
157 ematical methods. A cryptogram simply stated, is a message in a known language
158 encoded in an unknown script. A specific kind of cipher that has been thoroughly
159 studied is the single substitution cipher or cryptogram, where one sign in the script
160 corresponds to a single phoneme, syllable or alphabet. Newspapers and puzzles often
161 carry such cryptograms on their puzzle page. Typically the encoded message is a quote
162 by a famous person and is long enough to be uniquely deciphered. A homophonic cipher
163 is a variant of a cryptogram that assigns multiple signs per phoneme and may be used
164 to model scripts with allographs.
165 All available ciphertext for a given key constitutes a cryptogram regardless if the en-
166 cipherer sent the ciphertext in different pieces or logical units or at different times. This
167 may be trivially verified by picking random words from various websites, feeding them
168 into a cryptogram creator and solving on a cryptogram solver such as quipqiup.com.
169 This is because the cipher is a transformation function and all input to it is the
170 plaintext and all output is the cryptogram:
171 If M is the message, K the key, and E the enciphered message, or cryptogram, we have
172 E = f (M, K). In other words, E is a function of M and K. We prefer to think of this,
173 however, not as a function of two variables but as a (one parameter) family of operations
174 or transformations, and we write it E = Ti M The transformation Ti applied to message
175 M produces cryptogram E (Shannon, 1945).
176 There are two famous historical examples where separately captured messages have
177 been pooled together and used to break ciphers. First, the cipher of Mary, Queen
178 of Scots was cracked by treating 50+ of her enciphered letters as a single cryptogram
179 (George Lasry and Tomokiyo, 2023). Second, the Zodiac-340 cipher was mailed to three
180 separate newspapers and was cracked by analyzing them as a single cryptogram.
181 The captured ciphertext doesn’t need to be contiguous or even the same logical unit
182 of message, because every word of ciphertext reduces the equivocation of the key by
183 increasing the information we have about the legal adjacencies that may appear in the
184 plaintext. Language is a Markov process and the words and sentences in a language
185 form a graph representing the output of the process. These are termed “residue classes”
186 by Shannon. Residue classes are a subgraph of the Markov process. Possession of such
187 a subgraph reduces the equivocation dramatically. The more residue classes that have
188 been intercepted, the larger the subgraph and the smaller the equivocation. On average,
189 1/ρ residue classes of short words will enable the recovery of at least one sign value,
190 where ρ is the redundancy of the language. For most natural languages with ρ ≈ 0.7,

5
191 this works out to 2 to 3 short words.
192 The dictionary method using residue classes is deterministic and relatively robust
193 against homophonic ciphers as opposed to the frequency method, which is probabilistic
194 and ineffective in a script with many allographs. Residue classes are effectively regular
195 expressions in programming terms and make finding matches simpler. Attempts to use
196 frequency analysis on the Indus script did not yield any results other than the possibility
197 of some relation to the Brahmi script (Kak, 1988). The dictionary method of solving
198 cryptograms is also superior to frequency analysis since it’s immune to frequency drift
199 due to the passage of time (Moreno, 2005).

200 2.4. Shannon on Cryptograms

201 Claude Shannon’s groundbreaking A Mathematical Theory of Cryptography is a foun-
202 dational work of modern cryptography, signaling, data compression and information
203 theory (Shannon, 1945). This work addresses the mathematics and cryptanalysis of
204 cryptograms. According to Shannon, an unknown script such as the Bacon/Voynich
205 Manuscript may be modeled as a cryptogram (Bacon, 1401-1599). Our method to deci-
206 pher the Indus script essentially reuses Shannon’s cryptanalysis methods faithfully with
207 the help of modern programming languages.
208 A cryptogram may have many possible solutions, for example: the encrypted text
209 αβγ has ≈556 solutions, since any 3-letter word with no repeating letters is a solution.
210 Others like ǫǫλ only have a unique solution: EEL.
211 The question of the uniqueness of solutions to ciphers in general (as opposed to the
212 hand-crafted examples above) is useful to determine whether a decipherment is valid.
213 Shannon’s criteria for a cipher having a unique solution is determined by a quantity
214 termed “equivocation.”

215 2.4.1. Redundancy, Entropy and Equivocation

216 All natural languages have information content that is represented by symbols. Consider
217 the simple yes-or-no question: “Are you over the age of 18?” The answer to this question
218 may be represented as either of the words YES or NO. One does not need 5 total
219 symbols to denote this information, Y, N or even 1, 0 are sufficient. There are exactly
220 two alternative answers to this question and therefore it represents log2 2 = 1 bit of
221 information. The answer to whether the person is over the age of 18 therefore, can be
222 encoded as one binary digit. Similarly, the information regarding the day of the week
223 requires log2 7 ≈ 2.8 bits.
224 Natural languages use words and not bits and therefore use significantly more symbols
225 than necessary to transmit the same amount of information. Natural languages need to
226 use excess symbols because of dependencies and rules on how symbols are put together
227 to form words and sentences. This means that in natural languages, unlike optimized
228 binary notation, many combinations of phonemes (or letters or syllables) are illegal
229 words and are meaningless. For example, for the day of the week, if the first letter is
230 known to be T, then the next letter can only be {u, h} (for Tuesday and Thursday).
231 Other letters such as x, f are invalid as would give Tx, Tf. This inefficiency and excess
232 use of symbols is redundancy.
233 When we decipher a symbol or make a choice of value, we gain information about
234 the plaintext. The quantity of information gained when we make a choice is entropy.
235 Its counterpart, the available meaningful choices we can make about the unknown is
236 equivocation.

6
237 The equivocation for the day of the week, given that the first letter is T is two
238 alternatives {u, h} or 1 bit. If the first letter is W, then the equivocation is zero, since
239 only one alternative, Wednesday is possible. Natural languages have ≈70% redundancy
240 regardless of the language and conventional scripts used to write the language. This
241 redundancy can be measured experimentally and is very close to the compressibility of
242 the language. When the equivocation is zero, then by definition, the cryptogram has a
243 unique solution.

244 2.4.2. Unicity distance

245 The key space is the number of possible keys in a cipher. For a single substitution cipher,
246 the key space is the number of ways to rearrange the plaintext alphabet to create the
247 cryptogram alphabet. For an English single substitution cipher, the number of keys is
248 26 × 25 × ...2 × 1 = 26! The information contained in the key space is log2 (26!) ≈ 88.4
249 bits. A decipherment that reads much less than the information size of the key space is
250 dubious because the information of the deciphered text may exist in the equivocation of
251 the key space itself and may not be connected in any way to the ciphertext. A reliable
252 decipherment, therefore, must decipher plaintext of a length beyond the information
253 size of the key space. After a certain amount of plaintext is deciphered, equivocation
254 becomes zero and the cryptogram has a unique solution. This length is called the unicity
255 distance. The reliability of a decipherment that reads beyond the unicity distance may
256 be explained as the impossibility of information being created ex nihilo. If meaningful
257 information larger than the key space is extracted, then it must certainly be from the
258 ciphertext and not from the key.
259 A source language represents the plaintext, consisting of p symbols, each representing
260 a letter, syllable or phoneme. A homophonic cipher of N symbols, N > p, may be
261 constructed by mapping each ciphertext symbol to a plaintext symbol. Each ciphertext
262 symbol may be assigned to a plaintext symbol in p ways, thus carrying log2 p bits of
263 information per symbol.
264 The number of possible keys in a homophonic cipher with N ciphertext symbols
265 mapped to p plaintext symbols is < pN . Its key space is |K| ≈ log2 pN bits. The
266 redundancy of the plaintext language is denoted by ρ which is ≈ 0.7 for most natural
267 languages.
268 A large key space causes a large equivocation resulting in many possible meaningful
269 but false decipherments. The equivocation is also inversely proportional to the redun-
270 dancy of the language. Languages with low redundancy will have many more spurious
271 matches compared to languages with high redundancy. The length of the deciphered
272 plaintext that equals the information content of the key is known as the unicity distance
273 and is given by:

log2 pN N
d= =
ρ · log2 p ρ

274 A homophonic cipher may be viewed as insulating the ciphertext from the alphabet
275 of the source language as we can see from its unicity distance, which essentially depends
276 entirely on the ciphertext alphabet rather than the plaintext alphabet.
277 Of the total 417 signs, the 124 “ligatured” signs such as  and  and strings such
278 as  are simply read as if they are their component signs, they add no equivocation
279 and their count must be reduced from the ciphertext alphabet. Similarly, if the same
280 sign can be assigned to multiple phonemes, the count must be increased. Reusing 3̃0

7
281 signs of unaspirated phonemes for aspirated adds about 10%, as does reusing 2̃0 dental
282 signs for retroflex. Pooling all sibilants into 2̃0 signs increases the equivocation by 150%
283 for those signs. The two interchangeable signs  for a⇔e and  for a⇔o add two more
284 signs. The net effect would be adding (30+20)×0.1 + 20×1.5 + 2 = 37 signs. This
285 gives us a new symbol count of 417 − 124 + 37 = 330 and an effective unicity distance,
286 given the redundancy of Sanskrit at 0.7 (Aniket Anand and Jana, 2013):

log2 48330 330

= ≈ 471
0.7 · log2 48 0.7

287 Reading the longest 50 inscriptions of length 10 or longer covers this comfortably.
288 Note that this is for segmental decipherments of attested languages. Abjads would
289 require multiplying the signs by the number of vowels and decipherments into unknown
290 dialects are unfalsifiable.

291 2.5. Previous decipherments

292 It is impractical to enumerate and falsify the 100+ attempted decipherments of the
293 Indus script. Most are not serious attempts, but simply insightful observations and
294 preliminary attempts to read a dozen or so signs and as many inscriptions. These are
295 dubious for the same methodological flaws in the somewhat serious attempts that read
296 larger amounts of text as described below.

297 2.5.1. Logographic and Logosyllabic

298 Most attempted decipherments of Indus script are logographic or logosyllabic, avoiding
299 the need for syllables or phonemes to interlock into words. A majority of signs can be
300 claimed to be single words, most inscriptions are read as nouns with occasional generic
301 verbs. This enables short inscriptions to be read as adjective-noun or verb-adjective-
302 noun. The method starts by assuming the value of a sign and then tries to read a
303 meaning into a short inscription.
304 Parpola for example, considers the intersecting circles sign  to represent bangles
305 and three strokes sign  to represent hearth (Parpola, 1994). The reasoning for such
306 an assumption is unexplained. These two signs may as well represent nuts and three or
307 wheels and baby teeth or any thousands of competing concepts. The unicity distance
308 of such decipherments is at least the entire corpus. Occasionally, these signs occur in
309 meaningless patterns and need an alternate explanation. For example, when these two
310 signs are combined , the reading hearth bangles would be meaningless, so a new
311 overloaded interpretation is added: pregnancy bangles.
312 Incrementing the key space for every permutation of signs causes the key space to be
313 always larger than the entire corpus. Therefore, all decipherments using this method
314 are dubious. The information content of the readings is from the key and not from the
315 ciphertext. This is why there exist dozens of logographic and rebus decipherments that
316 read a lot of text but in ways completely unrelated to each other. A Rosetta stone
317 would make a logographic decipherment credible since it would double the information
318 content read and go beyond the unicity distance.
319 In some decipherments, some sign values are borrowed from other scripts. Signs are
320 assumed to have the same value as any similar sign in other scripts. Mahadevan for
321 example, assumes symbols similar to Egyptian hieroglyphs and Sumerian cuneiform

8
322 represent the same meaning in Indus script and then tries to connect it to some Indian
323 cultural or mythological context to decode a sign (Mahadevan, 2010). For example,
324 based on his analysis of M-1896  and the similarity of Indus sign  with Egyptian
325 hieroglyph O.49  meaning city, Iravatham Mahadevan speculated that Mohenjodaro’s
326 ancient name could be Kukkutarma, [kukkuta = Indian fowl] or ‘Cocks-city’ based on
327 interpreting the 1st CE Chola kingdom city Uraiyur in South India, as the city of the
328 cock. No evidence of chickens have been found in Mohenjodaro. Recent evidence shows
329 that chickens were domesticated in Thailand about 300 years after Indus Valley stopped
330 making seals (Peters et al, 2022). Mohenjodaro and Uraiyur which are separated by
331 2500 years and 2000 km are alluded to as perhaps connected for narrative support since
332 ‘cocks-city’ is a meaningless term not backed by any evidence.
333 It’s usually easy to find failures of this method for even short inscriptions. Using
334 the meaning of other signs from his decipherments, H-452  would be chicken-jar, M-
335 795  would be ruler-chicken-jar-bearer. Every combination of signs requires an
336 expansion of the keyspace to enable a meaningful reading. Logographic decipherments
337 even with borrowed sign values are untenable because the unicity distance is much
338 greater than the corpus size and the information content of the readings is essentially
339 from the key itself.

340 2.5.2. Borrowed starter set

341 Eclectic collation of sign values from many different scripts is the central method of
342 some syllabic and phonetic decipherments. Their advantage over logographic claims is
343 that they can use imported sign values as a starter set and discover the values of the
344 remainder using dictionary search and guesswork. S. R. Rao pioneered this technique
345 by claiming that similar-looking Semitic and Indus signs have the same values (Rao,
346 1980). He used the Semitic starter set supported simply by his assumption that Indus
347 signs eventually developed into Semitic signs.
348 The defect in this method is that the choice of a starter set is fairly arbitrary. Any two
349 scripts will have some signs that look similar or similar enough with completely different
350 sound values. Roughly 33% of both Linear-B and Caroline Island appear similar to
351 signs from Indus script but Indus inscriptions are unreadable if we borrow their values.
352 Simple signs with lines, circles, half circles or combinations are fairly common across
353 many scripts and similarities increase with the number of signs available. To illustrate
354 the defect in this method, we create a starter set for Latin based on some scripts that
355 have signs common with the Latin alphabet as in Table 1 and it would be incorrect to
356 use it to discover the values of the remaining Latin signs.

357 2.5.3. Segmental, syllabic, logographic and word-fragment alphabet

358 Some decipherments use arbitrary word fragments as sign values to fit as many in-
359 scriptions as possible. This in itself is not necessarily incorrect as long as sufficient
360 inscriptions can be read. Phonetic and syllabic decipherments would be expected to
361 have some grammar but nearly all prior decipherment claims typically lack grammar,
362 morphology, syntax, subject-verb agreement, plurals, conjugation, declension, preposi-
363 tions or any other grammatical elements. Often there are supplementary logographic
364 assignments to get around nonsensical readings. One way to justify the lack of grammar
365 is to claim that all Indus inscriptions are names. Short inscriptions can plausibly be
366 argued to be names since names outside sentences are essentially not declined. This
367 also means most long inscriptions often won’t be readable.

9
 Table 1. Eclectic decipherment of Latin alphabet from unrelated scripts

Source Symbol Value Source Symbol Value

Cherokee D a Inuktitut P ki
Cherokee W la Inuktitut L ma
Cherokee I qua Inuktitut U te
Cherokee L tle Inuktitut V pe
Cherokee T i Ethiopic O ’
Cherokee H mi Ethiopic M ṭ

368 Typically, this method starts with fitting popular inscriptions such as the Pashupati
369 seal and the Dholavira signboard. This force fitting produces garbled readings for many
370 other inscriptions. These are mitigated by assigning word fragments and word stems
371 to some of the remaining signs. Most inscriptions still won’t be readable in a known
372 language. The solution is to claim that the inscriptions are in a hitherto unknown
373 dialect. This dialect has no documented grammar, phonotactics or any known language
374 characteristics. The information content here is also entirely from the key, which is not
375 just in the sign values but a major portion of the key is the undefined grammar of the
376 language. Such decipherments therefore are dubious.
377 In Summary, every prior decipherment uses methods that create a key space that by
378 design is larger than the plaintext read and must be considered dubious.

379 2.6. Solving a cryptogram

380 Shannon’s method to solve a cipher:

381 (1) Determine the language.

382 (2) Determine the cipher (cryptogram for Indus script)
383 (3) Determine one or more letters of the key by checking patterns of occurrence (fre-
384 quency, variety of contact, doubles, reversals)
385 (4) Use the deciphered key portion as a consistency check by testing other parts of
386 the cryptogram.
387 (5) Repeat the step above until fully deciphered.

388 2.7. Determining the language

389 The first step to solving a cryptogram is to guess the language itself. If the language
390 is incorrect, the cryptogram will be unsolvable. Language is a Markov process and the
391 probabilities of symbol transitions are very different even for related languages, which
392 means that a large pattern set representing the corpus of one language cannot be read
393 as another even if symbol values are reassigned. A cryptogram whose plaintext is in one
394 language cannot be forcibly decrypted in another language due to the principle derived
395 by Shannon:
396 the amount of uncertainty we can introduce into the solution of the cryptogram cannot
397 be greater than the key size (Shannon, 1945).

10
398 2.8. Meluhhan language
399 Many languages have been proposed to be the language of IVC, the most credible
400 ones are some form of Dravidian (old Tamil) and Indo-Aryan (Sanskrit). Others such
401 as Sumerian-like language or a lost language have also been proposed without any
402 evidence. We may dismiss them as speculative.

403 2.8.1. Dravidian as the candidate

404 There are some reasons why Dravidian is unlikely to be the language of the Indus
405 Valley Civilization. As observed by many others, Dravidian has no words for the most
406 important IVC technologies, products or symbols but instead uses borrowed Middle
407 Indo-Aryan (Prakrit) words such as iṭṭika brick, gajja barley, swastika, paṭṭa[ṇa] city,
408 ūru city and there are missing words for the rest like the blackbuck, the unicorn, the
409 rhino. It is unlikely that Dravidians forgot the words for the important symbols and
410 technology they invented and continued to use till the present day while retaining their
411 language.
412 An analysis of Indus inscriptions also rules out Dravidian. Steve Bonta, a PhD in
413 linguistics in both Dravidian and Indo-European attempted to decipher the Indus script
414 as Dravidian for years until he noticed that the Indus inscriptions exhibit multi-stem
415 compounding, which is characteristic of classical Sanskrit and is not possible in Dra-
416 vidian languages. Dravidian compounds are two-word agglutinative constructs rather
417 than inflexed stem compounds. Consequently, he changed the target language of his
418 decipherment to Indo-Aryan (Bonta, 2023). Bonta’s observation has further implica-
419 tions that rule out all agglutinative languages as candidates for the Indus script such
420 as Dravidian, Sumerian, Elamite, Hattic, etc.
421 Like all agglutinative languages, Dravidian uses fixed affixes to a root to indicate
422 number, gender, noun cases etc. Agglutinative morphologies have a fixed order for
423 these affixes. Cinque argues for a universal order of these affixes whose order reverses
424 based on whether the language is head first or head final (Cinque, 1999). In Dravidian,
425 the plural affix comes before the locative for example. Switching these around results in
426 an illegal word. Also, these affixes by themselves (with rare exceptions) are not words
427 and affixes cannot exist freely without a root or stem. The Indus inscriptions show
428 the same strings of signs in initial medial and final positions, indicating that the Indus
429 script cannot be agglutinative but is rather a fusional language.
430 We can examine a few seals to see if there is any evidence of agglutinative morphology.
431 Consider the single sign inscriptions H-1497 , H-1546 , H-1514 , H-1462 . The
432 inscription H-246  cannot be read as the sign  having multiple affixes  since
433 affixes cannot be independent words. Moreover, if  is a stem and  is the stem
434 with three affixes, then we should not see the affix standalone in one inscription or its
435 signs inscribed in a different order as in H-894 .
436 It is clear from perusing the corpus that in many inscription pairs like B-9  and
437 Dholavira , common substrings such as  are not affixes, but rather stems that
438 are compounded. The terminal jar sign is likely to be a case marker. Thus we can
439 see the full word has a marker H-1953  but is removed for stem compounding in
440 M-1706  and placed terminally. Words that are compounds of 3 or more
441 stems like M-1706 are not possible in Dravidian but rather are characteristic of Sanskrit,
442 especially classical Sanskrit.
443 We can attempt to decipher the Indus script with Dravidian as the target language
444 using the dictionary method. Since proto-Dravidian has only been reconstructed to

11
445 around 800 words, it is likely to cause false negatives and therefore a Tamil dictionary
446 is more suited. We hit many dead ends with Tamil. Firstly, words with triple repeating
447 sequences are not present in Dravidian. So we would be unable to read inscriptions
448 like H-764 . Secondly, there are only a handful of words that would fit a doubled
449 sign inscription such as H-1182   and H-210 . These only match the words koko,
450 kūkū, māmā, tātā. This would assign one of ko, kū, mā, tā to all doubled signs. These
451 signs, however, also exist in permutations of themselves. Such inscriptions do not have
452 matching words in Tamil, so inscriptions such as H-2272  and H-372  would be
453 unreadable. This is an expected result if the Markov process of Tamil and the source
454 language of the Indus inscriptions produce dramatically different graphs and there is
455 no way to read one as the other. At this point, we can confidently rule out Dravidian
456 and indeed all agglutinative languages out of the running for the language of the Indus
457 script.

458 2.8.2. Sanskrit as the candidate

459 The question of how Sanskrit/Indo-Aryan could have existed in the Indian subcontinent
460 during the mature phase of the Indus civilization is something we need to address.
461 Heggarty’s Bayesian model indicates that Indo-Iranian and Indo-European started to
462 become audibly different around 4000 BCE and were markedly differentiated by 3500
463 BCE (Heggarty et al., 2023). This gives sufficient time for an old Indo-Aryan dialect
464 to be present in the Indus Valley for the bulk of the inscriptions which are dated post-
465 2600 BCE. The assumption that the Indus Valley was non-IE speaking is based on the
466 narrative that the Indus Valley could not have created Rigvedic poetry with horses and
467 chariots, which were allegedly unfamiliar steppe technologies. This narrative requires
468 us to ignore dozens of publications with evidence of horses from the earliest phases of
469 the Indus civilization and depiction of horse-driven chariots in the subcontinent from
470 the Chalcolithic era (Danino, 2006; Neumayer, 2020). The real issue is not ignoring
471 the evidence, but the presumption that one needs to possess horses to write poetry.
472 Ireland has used peacock imagery without the imposition of any claim that Indians
473 brought Gaelic to Ireland (Sheehan, 2009). And of course, there is poetry on unicorns
474 and dragons in the Rigveda without anyone needing to bring them. Evidence for trade
475 contact between the Indus Valley civilization and Sintashta is evident from the presence
476 of IVC cotton in Sintashta (Shishlina, Koryakova, and Orfinskaya, 2022). This alone
477 may be sufficient for Indus Valley civilization to create horse based poetry. The present
478 hypothesis on the spread of the Indo-European language family is an imperfect model,
479 while Indus inscriptions are actual attestations. If they conflict, the empirical method
480 requires the model to be adjusted and not the attestation.

481 2.8.3. Indus/Brahmi mixed inscriptions

482 Indus script signs continue to be embedded in later Brahmi scripts into the Gupta era
483 in both northern and southern India. An inscription from Vaishali  is equivalent
484 to , which is one of the most popular inscriptions that is attested in 40+ Indus
485 seals (Sinha and Roy, 1969). Every inscription in a mixed Indus/Brahmi script is in
486 the Sanskrit language, even in the southernmost and the oldest sites such as Keezhadi
487 in south India. We know the Indus script intermixed inscriptions in Tamil Nadu sites
488 are not Sanskrit words borrowed into Tamil but actual Sanskrit phrases because they
489 use signs such as the Brahmi 𑀱 ṣa which would have been changed to the Brahmi 𑀲 sa
490 on borrowing.

12
 Table 2. Names of Meluhhan persons and goods as recorded in Sumerian cuneiform

Artifact-Id Sumerian Meluhhan Name Meaning

CDLI 516366 sa6-ma-ar समर[MW] Samara confluence

CDLI 516366 na-na-sa3 न�नस� [MW] Nānas distinct
CDLI 525331 szu-i3-li2-su श� ल�श[ŚivaP] Shailesha Himalaya
CDLI 212982 sun2-zi-da स� �सद्ध[Bhag] Samsiddha accomplished
MS 2814a [szu?]-ib-ra श्वभ्र[MW] Śvabhra offerings pit
L 1426b me-luh-ha म� ल�c Melaḥ (person) union

CDLI 388265 mes म� ष� meṣī Dalbergia

CDLI 228643 dim3-dim3 �ड��डम ḍiṇḍima small drum
(Many)a me-luh-ha म� ल�� Melāḥ (country) settlements
a (George, 2003); b (Parpola, Parpola, and Brunswig
Jr, 1977); c (Devi, 1933); CDLI DB: (CDLI, 2023);

491 Based on Indus script signs embedded in various Tamil Brahmi and Sanskrit Brahmi
492 inscriptions, the most realistic candidate for the Indus language is Sanskrit.

493 2.8.4. Sanskrit morphology in the inscriptions

494 The attestation of triple symbols such as  would not fit many languages but could
495 be lit conjugation (somewhat similar to English perfect tense) of Sanskrit. Some roots
496 with a duplicated consonant become a triple consonant due to reduplication in lit. For
497 example, jajaja ‘I fought’ is first person singular perfect of √jaj ‘to fight.’ Similarly lalala
498 ‘I enjoyed’ from √lal, sasasa ‘I slept’ from √sas etc. In addition, the large number of
499 doubled signs could also be lit conjugation in plural second person such as rara ‘you all
500 gave’ from √rā or babha ‘you all shined’ from √bhā.

501 2.8.5. Sumerian references to the Indus Valley

502 Additional evidence comes from Sumerian inscriptions (Parpola, Parpola, and Brunswig
503 Jr, 1977). The Meluhhans are first mentioned in Mesopotamian texts in an inscription
504 of Sargon 2300 BCE. Sumerians traded timber, ivory, carnelian, water buffalo, gold
505 dust etc. with a land named “Meluhha.” Meluhhans built large sea-worthy ships capable
506 of carrying large animals across the sea. Only the Indus Valley civilization produced all
507 the products sourced from Meluhha and had extensive ocean trade. Meluhha therefore
508 is nearly unanimously believed to be how the Sumerians referred to the Indus Valley
509 civilization or one of its trading posts.
510 Meluhhans seem to have settled and intermarried into Sumer and some Meluhhans
511 adopted Sumerian names. There are some names, however, that are not Sumerian.
512 Most of these seem to be attested as proper names in Sanskrit literature. Meluhha
513 itself needs to be both the name of a person and the name of a country as shown in
514 Table 2. One of the names ending in ibra could represent many Sanskrit names such as
515 Vibhra, Atibhra etc. but unlikely in other candidates like Dravidian.
516 In addition, 90 identified words in Sumerian may be borrowed from Sanskrit (Vyas,
517 2020). This may be seen as sufficiently strong evidence to attempt decoding the Indus
518 language as Sanskrit. We use Monier-Williams abbreviations throughout the paper to

13
519 denote the source of attestation (Monier-Williams, 1899a). For example, Soma स�म[RV]
520 means “Soma as attested in Rigveda” and Samara समर[MW] means “Samara, attested
521 as a proper name in Monier-Williams dictionary.”

522 2.9. Determining one or more letters using occurrence patterns

523 The next step is to guess or decipher the first sign using unusual patterns in the text
524 that will enable us to read one or two syllabic words. We can rely on the fact that every
525 seal is one or more complete words and choose the shortest available seals to avoid the
526 issue of detecting word boundaries on longer inscriptions. With sufficient inscriptions
527 that only have a single common undeciphered sign, we can uniquely determine its value.
528 We can then repeat the steps and use the known signs to decipher the next unknown
529 sign.
530 All essential information for the cryptanalysis in a cryptogram is in the pattern of
531 occurrences of the symbols and not the symbols themselves. The plaintext EEL may
532 be enciphered as any of these equivalent cryptograms: BBA, XXC, GGF. Shannon calls
533 such a pattern underlying a set of equivalent cryptograms a residue class:
534 It is obvious in this case that these cryptograms are essentially equivalent. All that is of
535 importance in a simple substitution with random key is the pattern of letter repetitions,
536 the actual letters being dummy variables. Indeed we might dispense with them entirely
537 indicating the pattern of repetitions (Shannon, 1945).
538 Cryptanalysis of cryptograms depends entirely on the pattern of positions and repeti-
539 tions of signs in the ciphertext and does not depend on symbol shape, evolution history
540 of a sign, assumptions on cultural aspects, geography, etc. Indeed, because none of
541 these aspects are inputs, they can be credibly deduced as outputs once sufficient signs
542 are deciphered.
543 Patterns in the Indus corpus may be represented as regular expressions enabling
544 us to search in dictionaries and determine their values. Solving cryptograms using
545 regular expressions is a known art and we will only touch upon the techniques used for
546 deciphering the Indus script (Goibhniu, 2007).

547 2.10. The dictionary method of solving cryptograms

548 We illustrate the dictionary method with a simple example of an unknown script en-
549 coding text in the English language. There are 26 possible values for each symbol and
550 trying each value would require 26! possibilities, which is infeasible. If we can, however,
551 reduce the number of possibilities to solve an individual letter to two or three options,
552 then the script can be deciphered in under a hundred attempts.

553 2.10.1. Bootstrapping

554 Let us suppose we parse through the corpus of the enciphered text and isolate the
555 following words from the ciphertext. Whether these words were collected from one or
556 many messages does not affect the decipherment as long as the values assigned to signs
557 haven’t changed.
558 Words with repeated signs have low equivocation and are great for bootstrapping
559 our decipherment. An inscription pattern like /^(.)\1.$/ such as ǫǫλ matches the
560 word EEL uniquely giving us {E, L}. Other words may decode different signs due to
561 unique patterns such as P AP AY A and M ILLEN N IU M which can give a different

14
562 bootstrapping set.
563 Bootstrapping can be done even if there are no inscriptions with repeated signs, but
564 it takes a few more words to remove all the equivocation.

565 2.10.2. Further progress

566 Once we have a starting set of symbols, we can simply use them to find other sym-
567 bols. For example, to expand our decoded set of {ǫ, λ} = {E, L}, let us suppose we
568 locate another word λιǫ. We can substitute the values λ = L and ǫ = E and try to
569 find matches for ι using the regular expression /^L.E$/', which matches the words
570 {LEE, LIE, LY E}. For a homophonic cipher, E can be assigned to both ǫ and ι, so
571 we have 3 candidates for the value of ι = {E, I, Y }. We need to find another word that
572 reduces this set. Let us suppose we locate another inscription ιλλ, we can substitute
573 the values from our candidates in the new pattern /^(E|I|Y)LL$/ and we get a unique
574 result {ILL}. We have now decoded ι = I . Our decoded set expands to {E, L, I}. We
575 can continue to look for more words to help us decode more signs.

576 2.10.3. False negatives and false positives

577 It is pretty clear from the method that the accuracy of the dictionary and the correctness
578 of the regexes have a significant role in the correct decipherment of the script. Missing
579 words may give a false negative and the sign cannot be resolved and it would appear
580 we are at a dead end. At this point, we should check if the dictionary needs to be
581 augmented from the possible candidates. Often not all conjugations, compounded stems
582 or declined forms are available in dictionaries and false negatives should be expected.
583 If a conjugated verb, plural noun or gerund form is obvious, then we may simply add
584 it to the dictionary. If the sign being resolved is in the word stem and the other signs
585 are all known, then it is sufficient to simply match the stem itself. For example, if the
586 word KNOWING is missing in the dictionary, and we are trying to resolve KNO(.)ING,
587 it’s sufficient to just match KNO(.). This is how long compound words in inscriptions
588 can be matched.
589 On the flip side, a false negative can lead to a false positive. Suppose in our example,
590 the dictionary did not have the words LIE or LEE, the regex /^L.E$/' would match
591 only LYE and ι would be resolved to Y. This risk can be greatly mitigated by repeating
592 the resolution of every sign using many different subsets of inscriptions, all of which
593 should resolve to the same value. In addition, if there are plenty of attestations for
594 the sign, grammatically correct and meaningful values for all of them reassure us that
595 no false positive has been chosen. Incorrectly decoded signs will not match further
596 plaintext words downstream and quickly result in a dead end making this approach
597 self-correcting. For example, if ι is incorrectly assigned the value Y in our example,
598 many plaintext words like ILL and IS will fail to match despite occasional plaintext
599 words like DIE matching DYE. With sufficient distinct attestations, it becomes clear
600 which signs are incorrect. Signs which have very few attestations may never resolve
601 purely via dictionary check and a grammar check and a general theme consistency may
602 be required.

603 2.10.4. Translation and correctness

604 Translation is a post-decipherment activity and an occasional bad translation does not
605 invalidate the decipherment. Translating ancient inscriptions without spacing or punc-
606 tuation is challenging on its own. Consider an inscription: MAYBEFOREMAN. Purely

15
607 based on spacing adjustments, it could be read as MAY, BE FOREMAN or MAYBE
608 FORE, MAN or MAY BEFORE MAN. One way to address this is to look for con-
609 sistency in themes. If most of the inscriptions are about a workshop or factory, then
610 FOREMAN could be the right pick.
611 The abundance of choices on how to translate an inscription is generally a good sign
612 that the decipherment is nearly correct. If many signs are right but a few of the signs
613 are incorrect, they would typically force a break or the translation would not be in the
614 same themes as the rest of the translation. For example, if the letter E was incorrectly
615 deciphered as T, we would have MAYBTFORTMAN, forcing a word break near the
616 bad letter and we may attempt to read as MAY BT FORTMAN. Sometimes the bad
617 letter may be innocuous in one inscription, for example, if O was deciphered as I :
618 MAYBE FIREMAN. While this instance is grammatically correct, many others would
619 have breaks and the text would be nonsensical or incongruous with the general themes
620 of the rest of the inscriptions. These errant signs can be detected using the intersection
621 of the signs in all broken inscriptions and the decipherment process is repeated for such
622 signs with greater thoroughness.
623 If a decipherment is utterly incorrect, it won’t even be possible to make any words,
624 let alone morphologically consistent words and grammatically correct phrases that can
625 be read meaningfully. For example: mapagakajh cannot be made into a Sanskrit or
626 English phrase no matter how we dice it. This is because p or ph never follow m in
627 Sanskrit phonotactics, nor do words end in jh. Even plausible sounding short strings
628 like hena, hoja etc are non-words and it is unlikely that we get many matches because
629 the symbol space of a language is mostly filled with non-words.
630 A small subset of short inscriptions may be read as an abjad and with wide latitude
631 for word forms, ignoring morphology or syntax with a partially correct decipherment.
632 The grammatically correct reading of the longest inscriptions thus becomes a litmus
633 test for decipherment accuracy.

634 2.10.5. Adjustments for Sanskrit in an ancient script

635 From the Meluhhan words and names attested in Sumerian inscriptions, it appears that
636 Meluhhans spoke some form of post-Vedic Sanskrit, but one that is too early to be
637 termed classical Sanskrit. This enables us to use a downloadable Sanskrit dictionary
638 with some caveats. Most dictionaries do not have compound, declined, conjugated or
639 affixed forms so we need to add these forms as needed. Many words like root forms,
640 grammar terminology and those created by medieval lexicographers should be removed.
641 A few words may be potentially lost over the millennia, and we may find some inscrip-
642 tions unreadable despite deciphering the script accurately.
643 Indus script is a bronze age script and we have no idea how much fidelity an inscription
644 has to the actual spoken language. Based on very early scripts, our derivation should
645 handle a script that is unlikely to have a virāma (i.e., handle a consonant without an
646 assumed a vowel). We need to accommodate for the possibility of sign reuse among
647 dentals and retroflexes, aspirated and unaspirated and possibly voiced and unvoiced,
648 similar to later Tamil Brahmi. Doubled consonants may also be written as a single
649 sign(i.e., datta written as data).
650 We adjust for these by flattening sibilants together and also dentals with retroflexes
651 and aspirated with unaspirated. Doubled consonants are rare enough that we can only
652 examine them when we run into a dead end. This enables us to use simpler regexes
653 and reduces the chances of user error in derivation. The capture group (.) captures a
654 single consonant or a single vowel. To capture either a syllable or consonant without

16
655 a following vowel, we would need to use (..?) as the capture group. The capture
656 group (..?)\1 would also capture reduplicated consonants but it’s an unwarranted
657 complexity. The capture group (.+) may be used to capture conjunct signs like .

658 2.11. Sign classification

659 Although we have used Mahadevan’s 417 signs as valid for a decipherment-neutral metric
660 of correctness, it is by no means authoritative. Indus script is a hand-carved writing
661 system and not a typeset or a tool-based script (unlike say, cuneiform). Therefore, just
662 as every handwriting sample of the same letter is unique, no two instances of any Indus
663 sign are exactly alike. There is plenty of room for stylistic and regional variants and
664 present classifications are subjective for the most part.

665 2.11.1. Signs and Glyphs

666 A sign is an idea of visual representation of a member of the script. A glyph is a par-
667 ticular realization of a sign and it’s normal to have multiple glyphs that represent the
668 same sign. For example, A A A A are all different realizations of the same sign in the
669 Latin alphabet. To be properly deciphered, the idea of the sign must be articulated
670 clearly, for example, The first letter A of the Latin alphabet is two vertical lines joined
671 at the top and separate at the bottom with a horizontal connection midway. This spec-
672 ification is sufficient to read an enormous array of handwriting and stylized fonts. For
673 pictorial signs, something like the phoneme D is a pictograph of a dog may be sufficient.
674 Regardless of the breed and orientation of the dog pictured, its intent is unambiguous.
675 A script made of pictorial glyphs is unlikely to have different signs based on tiny
676 variations such as “fish with two fins” vs “fish with 4 fins.” They are likely to represent
677 the same idea “a fish” and therefore different glyphs of the same sign. A rarely occurring
678 sign that is only slightly different from a frequently occurring sign is likely to be a
679 stylistic variant rather than an unrelated sign. Usually, there are unique characteristics
680 for variants of the same sign that help us correctly identify variants. For example, the
681 jar sign is unique for having zero to four small numeric strokes within it      
682  . All renditions of this sign are treated as variants. For simplicity, we use the most
683 common rendition in the paper. For example, any of    will be denoted as  in
684 the derivation section. While variants of most signs are quite apparent, we present a
685 concise list of the most expansive variants in table 3.
686 Rare mirrored signs like   are likely an artifact of carving for seal impressions.
687 Bracketed signs    are likely to be word fragments consisting of three or more
688 signs rather than a single unrelated sign simply because the brackets and numerics occur
689 independently for example in M-2069 . These presumptions are validated if we can
690 read nearly all bracketed signs as a string of their constituents. Similarly, if conjunct
691 signs like    are indeed digraphs, then nearly all digraphs should be readable as
692 their constituent signs. Gradual abstractions of complex signs such as     should
693 independently be deciphered to the same value.

694 2.11.2. Stylistic and geographic variants

695 While an ornate rendition of the first sign of an inscription may be used, different glyphs
696 are often stylistic variations that may rarely be mixed in an inscription. For example,
697 Upper Case Is Rarely Mixed With Lower Case except in titles and italics are rarely
698 mixed with bold. The scarcity of mixed styles gives us a proof point to verify that

17
 Table 3. Six extensively stylized signs

Sign name Stylized variants

Barley merchant 

Spider/Insect 
Spider/Insect 
Horse/Equid 
Comb 
Leaf 
Chariot 

Table 4. Discovery of distinct signs and allographs

Glyphs Inscription Matches Result

H-1850  bhaṭṭi

 vs  Distinct
Dholavira  mānana māyya mālla

M-874  rayi yayi

 vs  Allographs of ra
H-1182  śaraṇi

699 certain glyphs are indeed variants of the same sign if their values are identical.
700 When individual variants have sufficient attestations, we may prove each variant
701 individually for example, all 3 variants   may be independently proven to have
702 the value क ka. Where sufficient attestations don’t exist we can narrow down the
703 possibilities to a handful and attempt to read all the inscriptions meaningfully. For
704 example, by applying the regex method to identify  we do not get a unique value, but
705 narrow it down to र ra or व va. Based on sign similarity and readability in inscriptions,
706 we can confirm that the sign  is a variant of  र ra.
707 This method is not only helpful in identifying variants, but it is also self-correcting
708 and an incorrect identification will be easy to identify:
709 (1) two signs which are different are assumed to be the same
710 (2) two signs that are the same are assumed to be different
711 In the first case, regular expression search yields two disjoint sets of inscriptions with
712 two different phonetic values. For example, if we presume the signs  and  are variants
713 of the same sign, An attempt to narrow the matches from the first set using a member
714 of the second set will yield a null set. This immediately alerts us to the fact that the
715 signs are indeed distinct.
716 In the second case, both sets will yield the same phonetic value. In fact, this is how
717 we discover allographs as shown in Table 4.
718 While an optimal reachability graph can be constructed by analyzing the inscriptions,
719 we can use the heuristic of solving signs in descending order of frequency with a few
720 detours as needed.

18
721 2.12. The first sign
722 The first sign we decipher is the jar sign  representing an. It also represents a variety
723 of nasal sounds including anusvara, which is a post-vocalic nasal sound. From the
724 dictionary itself, a regex for H-764B  we get a single match ananan, representing
725 ananaṃ, which is an accusative of anana. Technically, there are other possible matches
726 like jajaja, lalala which are legitimate matches. The seal Dmd-1  single jar sign
727 matches ana and eliminates other alternatives like ja and la. Finally, many inscriptions
728 with terminal jar signs representing the accusative ending aṃ resolve the value of the
729 sign.
730 After we decipher the first sign, we simply look for other inscriptions that enable
731 us to decipher more signs by substituting the value of the first sign. The vertical bar
732 sign  is decipherable with just the jar sign and we keep repeating the steps till all
733 signs are resolved. We continue the process roughly going in the descending order of
734 frequency of signs, to optimize the probability of finding sufficient attestations needed
735 for deciphering the sign. The complete derivation may be followed step by step in
736 Section 8. A computer program that partially reproduces the derivation is available in
737 the supplementary section.

738 2.13. Formalization

739 Let G = {g0 , g1 , g2 ...} be the set of symbols whose sound value is known.
740 Let X = {x0 , x1 , x2 ...} be the set of all symbols.
741 Let S = {s0 , s1 , s2 ...} be the set of inscriptions where si is a vector from symbols in
742 X

743 (1) Let two inscriptions sa = gi xa gk xc ... and sb = gj xa gl xf ...

744 (2) Let set of symbol matches Gxa = /gi (.+)gk . + / ∩ /gj (.+)gl . + /
745 (3) Repeat intersection with new inscription above until Gxa is of unit length
746 (4) x a = G xa
747 (5) G = G ∪ xa
748 (6) Repeat until G = X

19
749 3. Results

750 Indus script is made of 76 allographs representing the most common signs shown in
751 table 6. The complete derivation can be traced in section 8. A compact one-pager is
752 given in table 5. Long inscriptions and inscriptions with conjunct signs are listed in
753 sections 5 and 6 respectively. Some signs have variants that mark reading direction or
754 word boundaries and they are listed in table 10. Consonant clusters using syncope are
755 shown in table 9. Table 7 shows Indus signs standardized into Brahmi.
Table 5. Indus script phoneme glyphs

Phoneme Glyphs
अ a              
आ ā 
      
इ i 
ई ī 
उ ū 
ए e 
ओ o 
अन� अ� an aṃ 
अस� अ� as aḥ 

कख k kh         
गघ g gh 
चछ c ch 
जझ j jh    
तथटठ t th ṭ ṭh  


दधडढ d dh ḍ ḍh                   
न n           
पफ p ph   
बभ b bh 
म m 
य y 
र r 
ल l 
व v     
सशषह sśṣh                         

20
756 3.1. Sign values and their reconstructed names

Phoneme Reconstructed Sign Name Sign Glyphs

अ a अय�ग ayuga one[VarBṛS]  1

आय� āyu man[RV]  2

अग aga mountain[Kirat]  3

अङ्क aṅka curve/hook[RV]      4

आज�न ājani stick[AV]  5

अजशृङ्ग� ajaśṛṅgī goat’s horn[AV]  6

आ ā repeated अ a 
       7

इ i इष�क� iṣīkā stalk of grass[AV]  8

ई ī doubled इ  9

उ ū उ��म udyāma rope[TS]  10

उदप�न udapāna well[MBh]  11

उप�नहन� upanihan hammer[ŚBr]  12

ए e एक eka one[RV]  13

ओ o ओपश opaśa pillar[RV]  14

अन� अ� अ� श� aṃśu soma drink[ŚBr]  15

an aṃ अ� श� aṃśu lamp[MaitrUp]  16

अस� अ� अष्टप�द aṣṭapāda eight legged[MBh] (spider)  17

as aḥ अष्टन� aṣṭan eight[RV]  18

कख क�तम� kṛtam dice[ŚBr]      19

k kh क�तम� kṛtam axe[Cutting √क�त� √kṛt]  20

ख�त� khātṛ digger/pickaxe[Cāṇ]  21

क�ष kṛṣa ploughshare[Gal]  22

गघ ग�धन gādhana arrow[Hariv]  23

g gh

चछ चत�र� catur four[RV]  24

c ch छत्त्र chattra mushroom[CarakaS]  25

जझ झर jhara waterfall[Prab]  26

j jh झञ्झ�न� jhañjān rain and wind[KāśīKh]  27

तथटठ तण्ड� ल taṇḍula rice plant[AV]  28

t th ṭ ṭh त�ड� ल tāḍula fighter; beater[Uṇvṛ]  29

त�ड्य tāḍya drum[Mn]  30

त�ल tāla small cymbal[BhP] 



31

त�� tardū wooden ladles[L]  32

तद� tarda Indian blackbird[AV]  33

त्र tra three[RV]  34

दधडढ धन्वन� dhanvan bow[RV]  35

d dh ḍ ḍh ध�न dhāna receptacle[RV]  36

दन्त danta teeth[RV]  37

ध�नक�� dhānakāḥ coins[Car]      38

21
Phoneme Reconstructed Sign Name Sign Glyphs
न n न�ल nāla mat of reeds[BhP]  39

न�ल nāla stalk[MBh]  40

न�ल�क� nālīka arrow[MBh]  41

नल nala reed[Bhp]  42

नवन� navan nine[RV]  43

पफ पञ्चन� pañcan five[RV]   44

p ph पञ्चन� pañcan hand[RV]  45

बभ भक्षपत्त्र� bhakṣapattrī edible[RV] leaf[RV]  46

b bh भक्षत्र bhakṣatra oven[भक्ष bhakṣa:RV]  47

म m मत्स्य matsya fish[RV]  48

मत्य matya club[AV]  49

मत्य matya harrow[TS]  50

मत्त matta elephant[R]  51

मन्थ mantha churning stick[MBh]  52

मन्थ mantha fire sticks[MBh]  53

मन्द�र mandāra a flower[MBh]  54

म�न्दर mandira dwelling[MBh]  55

मय maya horse[VS]  56

म�क्ष mṛkṣa comb[RV]  57

य y य�व�ष्ट yaviṣṭhi barley[RV] seeker[RV]  58

यव yava barley[RV]  59

य�ष्ट yaṣṭi pearl necklace[VarBṛS]  60

य�ष्ट yaṣṭi twig[Hariv]  61

र r रथव� ratharvī split snake[AV]  62

रथ��रन� rathārin chariot[RV] wheel[BhP]  63

रथ ratha chariot[RV]  64

रथद�रु rathadāru chariot wood[Pāṇ]  65

ल l लत� latā creeper[MBh]  66

व v वत� vartī lamp wick[MBh]  67

�व vi two[RV]  68

व�टक vīṭaka betel leaf/nut[Pañcad]  69

वट� vaṭī banyan tree[MBh]  70

व�त्र vātra loom [√व� + ष्ट्रन� ]  71

सशषह श�ख�र śākhāra squirrel; branch mover[RV]  72

sśṣh श�क्र śukra seed[RV]  73

�शखर śikhara mountaintop[MBh]  74

�शख� śikhā peacock crest[MBh]  75

श्य� न śyena falcon[RV]  76

श�ण śāṇa weight of four[Hariv]  77

षण� ṣaṇ six[RV]  78

सप्तन� saptan seven[RV]  79

स�प�न sopāna ladder[MBh]  80

81

22
Table 8. Numeric and pictorial line stroke signs

Numeric from digits 1-10 Pictorial

               
अ व त च प ष स अस� न द अ स ज न ष ज
a v t c p ṣ s as n d a s j n ṣ j

Table 9. Consonant clusters using schwa and double syncope

Seal-Id Inscription Sanskrit Transliteration Conjunct

H-43  शक्र śakra क्र kra

H-152a  क्षय kṣaya क्ष kṣa
M-43a  मज्ज वम�हन� majja vamāhan ज्ज jja
M-1316  स�ह वल्लभ sāha vallabha ल्ल lla
बड�द baḍāda

Table 7. Indus signs standardized into Brahmi

अ a इ i उ u ए e ओ o
𑀅  𑁦𑁇𑁦  𑀉  𑀏  𑀑 
क k ख kh ग g घ gh ङ ṅ
𑀓  𑀔  𑀕  𑀖  𑀗 
च c छ ch ज j झ jh ञ ñ
𑀘  𑀙   𑀝𑀝  𑀜 
𑀚

ट ṭ ठ ṭh ड ḍ ढ ḍh ण ṇ
𑀝 𑀞  𑀟  𑀠  𑀡 


त t थ th द d ध dh न n
𑀢  𑀣  𑀤  𑀥  𑀦 
प p फ ph ब b भ bh म m
𑀧  𑀨  𑀩  𑀪  𑀫 
य y र r Brahmi ल l व v
𑀬  𑀭  vs 𑀮  𑀯 
श ś ष ṣ Indus स s ह h
𑀰  𑀱  𑀲  𑀳 

23
 Table 10. Directional and word-boundary variants

Sign Marked Variants Sign Marked Variants

अ a     न n  
अ a   ब b  
अन� an   म m  
अन� an   म m  
अस� as   य y  
द d   र r  
द d   र r  

757 4. Discussion

758 Indus script is a segmental script that may be described as proto-abugida. The major
759 difference from Brahmi is that retroflexes use the same signs as dentals and aspirated
760 and unaspirated stops use the same signs. Signs have a default vowel of अ a unless
761 overridden by an immediately following vowel sign. No diacritics are used.
762 Many signs have stylistic variations that have no phonetic distinctions. While signs
763 are distinct in most cases, occasional ambiguities do exist, which are discussed below.
764 Signs may be classed into several groups based on characteristics specific to them.

765 4.1. The language

766 Many words from classical literature such as the Ramayaṇa and Mahābhārata are at-
767 tested, but the meanings of many archaic words are from the Vedic corpus. Chain
768 compounds are also a feature of classical Sanskrit. Although the grammar is essentially
769 the same, we refer to the Indus language as post-Vedic, to avoid confusion with the
770 classical literature that began with Pāṇini’s Aṣṭādhyayī.

771 4.2. The inscription content

772 The contents of the Indus inscriptions are similar to Janapada coins and Gupta-era
773 seals that they resemble(Azad, 2020). Long seals seem to use a concise Vedic con-
774 cept as a motto or slogan. This tradition is attested at least from Gupta times and
775 carries on to the present day. For example, a Gupta era seal reads śaṃkara-vara yānāb-
776 hyaḥ(IndMuseum, 5th Century). This translates to “Shiva’s blessing for the journeys”
777 which is a concise form of Rig-Veda 7.35.15. This could be a prayer for a journey of trade
778 goods. Another reads Śivam brihaspate “[Our] welfare, O bṛhaspati”(MetMuseum, 6th
779 CE). This is a concise form of Rig-Veda 1.89.6. The tradition of using Rigvedic for-
780 mulas as prayers and mottos continues to the present day in many organizations. The
781 government of India’s satyameva jayate “Truth alone triumphs” from the Chhāndogya
782 Upanishad, is the most famous. In addition, the various state and local governments of
783 India, Thailand and Indonesia continue to use concise forms of Vedic verses as mottos.
784 The first few signs often invoke a deity. Most of these are of the meaning of destroyer
785 or roarer. The Vedic god Rudra is both Roarer and Destroyer. This may be because the
786 root ru means both roar and kill. Most names are simply constructed by the -a affix to
787 the root, which creates the agent-noun from the verb-root (i.e., roarer from roar). In

24
 Table 11. Provisional readings of mixed IVC/Brahmi inscriptions from subsection 1.1

Source Inscription Reading Translation

1 Keezhadi 𑀘𑀸𑀧𑀬 च�पयशम� renowned archer

600 BCE cāpayaśam
2 Marungur 𑀅𑀫𑀳 अमहस� the mighty one
200 BCE amahas
3 Annakodai 𑀓𑀸𑀯𑀸𑀢 रर क�व�त� given by Kāva
300 BCE rara kāvāt (a Sāman)
4 Vietnam Gold foil 𑀯𑀭𑀸  वर�ह Varaha avatar
100CE varāha
5 Copper Hoard 𑀰𑀁𑀔𑀸𑀢 𑀓𑀺𑀫 श� ख�त� �क� what emanated
1800 BCE 𑀞𑀟𑀬 रठ� डय from the conch?
śaṅkhāt kiṃ
raṭaṇ ḍaya
6 Bet Dwarka 𑀔𑀸 𑀫 म�रख� ईश O Ocean expanse!
(R to L) <1000 BCE mīrakhā īśa O Lord!
7 Bijar-Ghamchoghai  𑀬  रूर रय�� O fire, the rivers
(R to L) undated   मम मठम� are my home
rūra rayāḥ
mama maṭham

788 Paṇinian grammar, this is called the अच� ac affix. Although Pāṇini mentions this rule
789 to apply only to a specific set of roots (pacādi roots), in practice this affix is universal
790 and can be applied to any root as evidenced in the oldest books of Rigveda [च�द� + अच� =
791 च�द cud + ac = coda RV 5.61.3].
792 The intended meanings of short seals are harder to determine with precision. For
793 example, the inscription H-101a  parṇa has an ordinary meaning of leaf, wing,
794 or feather, but is also a proper name of a people, a place, a teacher or a particular
795 tree. For the sake of conciseness, we just translate it simply as “leaf.” Most short seals
796 resemble the contents of the seals from the Gupta period and therefore are likely to be
797 names or membership tokens(Azad, 2020). Determining the actual intended meaning
798 of any inscription is beyond the scope of this work. For determining the accuracy of
799 the decipherment, it only matters that the inscription has been read grammatically
800 correctly.
801 The only properties of a language of interest in cryptography are statistical properties.
802 What are the frequencies of the various letters, of different digrams (pairs of letters), tri-
803 grams, words, phrases, etc.? What is the probability that a given word occurs in a certain
804 message? The “meaning” of a message has significance only in its influence on these prob-
805 abilities. For our purposes, all other properties of language can be omitted.(Shannon,
806 1945)

25
 Table 12. Provisional readings for non-Sanskrit seals from Mesopotamia and Dilmun(Laursen, 2010)

Source Inscription Reading Translation Language

Karzakan  an-an Sky-heaven Sumerian

Karzakan  a’aš am sign of the wild bull Sumerian
Karzakan  a’aš aga sign of the crown Sumerian
Mesopotamia?  kaš an-an decision of the sky god Sumerian
Mesopotamia?  maa-a boat labor Sumerian
Saar  ḫamu-a for raft Old Akkadian
Susa  karānu wine Akkadian
Ur  kamānu cumin Old Akkadian

Table 13. Validated analyses from past decipherments

Authors Finding

Mahadevana Primary direction is Right-to-left

Parpolab Signary contains allographs
Rajesh Rao et. al.c non-IVC language inscriptions in West Asia
S. R. Raod Language is Indo-Aryan
Huntere Indus script and Brahmi are related
Sullivanf Brahmi signs in Indus script have identical values
Bontag Semiotics suggest Indo-Aryan language
Heggarty et. al.h Indo-Aryan and Iranic separated by 3000 BCE
a (Mahadevan, 1977);b (Parpola, 1994); c (Rao et al., 2009); d (Rao, 1980);
e (Hunter, 1934); f (Sullivan, 2011); g (Bonta, 2023); h (Heggarty et al., 2023);

807 4.3. Validations

808 Although the results of this decipherment are markedly different from prior decipher-
809 ments, we validate many important findings of prior decipherments. These are listed in
810 table 13.

811 4.4. Allographs

812 The script has 76 allographs for the most common signs with a median of three allo-
813 graphs per phoneme shown in Table 6. These allographs seem to have appeared due to
814 mispronunciations when the name of the sign was transmitted orally. This may have
815 happened sometime after the Harappan period 3A when speakers of other Indo-Aryan
816 dialects began using the script as part of the Indus Valley integration phase. This is
817 analogous to the effect of Prakrit on the writing of the Maurya and Gupta eras.
818 A comprehensive allograph chart is in table 6. An average of a little over 4 variants
819 per allograph for 76 allographs explains the approximately 125+ signs.

26
820 4.5. Line strokes and Numeric signs
821 A numeric sign does not encode an actual number, but rather the first syllable of its
822 name. The sound values are shown in Table 8. A sign with eleven strokes does not exist,
823 suggesting the Indus civilization used a decimal system. The name for a single numeric
824 stroke, अय�ग ayuga, represents the अ a vowel. अय�ग ayuga may have mutated into अइक
825 aika which may have been used as the diphthong अइ ai. The word एक eka evidently
826 evolved from अय�ग ayuga via अइक aika. The other numeric signs are self-explanatory,
827 each representing the initial consonant of its name.
828 Non-numeric signs seem to have been invented earlier than numeric signs, as evi-
829 denced by workarounds to comply with non-numeric signs. Evidence for this is that
830 other अ a signs can be doubled to form आ ā but not the  vertical stroke since it would
831 be read as  स sa and inscriptions work around this by inserting  between the two to
832 create . The sign  ja often written as  representing rain, is not a numeric sign.
833 Its allograph  is not to be confused with the numeric sign  read as त ta.

834 4.6. Sign disambiguation

835 Numeric signs in general take all strokes as a single sign. Spacing, stroke size differences
836 and incline differences are used to distinguish two adjacent numeric signs or signs made
837 of simple line strokes.
838 For example, the sign M-734 reads    वरण-जज varaṇa-jaja, “chosen[Uṇādi] war-
839 rior[Śiś].” The clear space between   is unambiguously read as two  ज ja signs, rather
840 than a single  sibilant sign, which would need to be read as वण�स� varṇas, “color[MBh].”
841 Seal M-1822  uses incline differences to distinguish   स� sā to avoid being read as
842  ज ja.
843 When strokes are arranged in rows, if the upper row has an equal or higher number
844 of strokes than the lower row, then the sign is read as representing the total number of
845 strokes. Inscriptions use spacing to ensure clear separation between distinct signs.
846 For example, the sign M-1904 reads  जज jaja, “warrior.” The three-stroke sign
847 representing the first syllable is marked as distinct from the other by horizontal and
848 vertical spacing. If there was no clear spacing, the seal would look like  and be read
849 as a sibilant.
850 When the upper row has fewer strokes than the lower row, spacing or other indications
851 of separateness become important for an accurate reading. For example, M-948 reads
852  अववज्य avavajya “off[RV] journey[√वज� ]”, rather than अवपय avapaya “desiring[RV]
853 water[RV]” due to slight offset of two rows of strokes. While signs with five or more
854 strokes can appear in one or two rows, signs with two and four always appear in a
855 single row. If they appear in two rows, they are distinctly different signs with distinctly
856 different sound values. Note the differences between  व va,  ज ja,  च ca versus  न
857 na,  त ta,  श śa.

858 4.7. Vowels

859 A vowel sign overrides the default अ a vowel of the preceding sign just as in abugida
860 unless it’s another a, which lengthens to आ ā. In practice, a leading अ a can be either
861 अ a or आ ā. Repeated vowel signs make a long vowel. The only long vowels attested
862 are आ and ई. Other vowels such as ऊ simply use the same sign  to indicate both long
863 and short vowels. ऋ ṛ uses signs for र r and presumably so do the unattested ॠ ṝ ऌ
864 ḷ. The signs for  अन� an and  अस� as represent अ� ṃ and अ� ḥ respectively and have

27
865 vowel length flexibility, the rest are abugida and always read with default vowel a or
866 without any vowel and conjunct with the next sign.

867 4.8. Diphthongs

868 The sequence अइ a-i exists as a diphthong in some seals indicating that is possibly used
869 for ऐ ai as in H-1056  जमन� � jamanaiḥ [√जम� + ल्य�ट� n.ins.pl of noun from √jam].
870 An equivalent diphthong for औ au is not attested but may be reconstructed as .

871 4.9. Base consonants

872 A base consonant sign represents an ordinary consonant with the default vowel अ a.
873 Retroflex ṭ ḍ and dentals t d use the same signs, similar to the usage of the Latin
874 alphabet to write Sanskrit. Aspirated kh gh ch jh th dh ph bh use the corresponding
875 unaspirated k g c j t d p b signs.

876 4.10. Sibilants and nasals

877 All signs for श ष स ह ś ṣ s h are interchangeable including अस� as signs. These are
878 also used as visarga ḥ where needed. The usage of the as signs for h ḥ may reflect a
879 grammatical feature of Sanskrit where a nominative affix s becomes ḥ in sandhi. The
880  sign is also used nasal consonants न n, ण ṇ and ङ ṅ. This interchangeability of
881 retroflex with dentals, aspirated with unaspirated, the sibilants among themselves and
882 the same with the nasals may be additional evidence of the presence of multiple Indo-
883 Aryan dialects in the Indus Valley civilization. We see the same effect in the Sanskrit
884 words attested during historic times due to the presence of various Prakrit dialects.
885 For example, Sanskrit स� न sena becomes Prakrit ष� ण ṣeṇa, रज्ज rajja is written as रझ
886 rajha and so on. To Prakrit speakers, these signs are interchangeable in a script. More
887 examples are shown in table 15.
888 The  variant seems to be preferred for ञ ñ but in principle, all jar-like signs are
889 interchangeable. Ambiguities that may be caused by the flexible vowel length of jar signs
890 are avoided by adding an अ a right after. For example, L-218  would be read as
891 अनवस anavasa, Rivedic[6.66.7] word meaning “unstopping” describing the unstoppable
892 march of the storm gods, the Maruts, rather than a fairly meaningless अण्वस aṇva-sa
893 “interstice bestowing.”

894 4.11. Conjunct signs

895 A conjunct sign is created by combining two (or rarely more) base signs and represents
896 the sounds of the combined signs. The most common being the signs for अम ama,
897 which can take a variety of forms by vertically assembling a rotated curve sign  अ a
898 and one of the म m signs, giving many possible conjuncts,  roofed-fish sign being the
899 most common. In addition to these, there are over 100 rarely used conjuncts that are
900 deciphered in Section 6.
901 Conjuncts appear to be artifacts of limited space. One type of conjuncts is constituted
902 of two base signs that appear to be ligatured because they touch each other due to
903 crowding caused by lack of space. Extremely rare occurrences of conjunct signs clarify
904 that a conjunct sign is simply two or more normal signs that just happened to touch due
905 to crowding. They may be simply read as if they are separate. Often signs are missed

28
 Table 14. Directional markers on jar signs

Seal-Id Inscription Sanskrit Transliteration Direction

H-1711  क�न्तद�न kāntadāna ←

M-1353  अन्नद�न annadāna ←
M-1304  अनवरम� anavaram ←
H-1801  अनवरम� anavaram ←
M-862  अनरम� anaram ←
M-420  अन�चर anācara ←
M-1822  अन�ष� anūṣā ←
M-1700  अन्ध�� andhaiḥ ←

Ns-86  अन� anu →

M-1336  आनवम� ānavam →

M-2062  line1: चर cara ←

 line2: अ� श aṃśa →

M-378  वनम� अपर vanam apara word boundary

H-454  अनम� वरम� anam varam word boundary

906 during the initial carving and later squeezed if space permits as in C-8 . These
907 are read as if they were normal-sized. If there is insufficient space to squeeze in the
908 sign, then a narrow sign like  or  may be rotated and placed above the next character
909 as in   . These are simply read as two separate signs with the top sign followed by
910 the bottom within the flow of the text. Occasionally, there is insufficient space to insert
911 a character in the right location, so a proper conjunct is made, for example,  ji in
912 M-409 . Such conjuncts appear at the ends of inscriptions where there is no space
913 to insert the missed character, so the missed sign is conjunct to its nearest one and may
914 be read with the (normal sized) base character first and the attached sign following
915 it:  is read as . Exceptions can be made based on words that are attested
916 in other inscriptions like  dhakka but these are rare. The vowel conjuncts may have
917 evolved into Brahmi diacritics. For example, the Indus conjunct  may have become
918 Brahmi 𑀥𑀻 dhī.

919 4.12. Consonant clusters

920 Clustered consonants are simply written adjacent but are pronounced with a schwa
921 syncope. Doubled consonants are written as a single consonant. Examples are shown
922 in Table 9.

923 4.13. Directional markers

924 While the preferred direction of writing is right-to-left as discussed in section 2.1, occa-
925 sional left-to-right and boustrophedons are also attested. Egyptian hieroglyphs and the
926 Phonecian scripts simply mirror the signs when the direction changes, so boustrophe-
927 dons are simple to detect. Since the vast majority of Indus script signs are symmetric

29
928 and sometimes even the asymmetric signs are inscribed mirrored horizontally, there is
929 a chance that the inscriptions are read in the wrong direction. For large inscriptions,
930 this is usually not a problem, since they would be unreadable the wrong way. However,
931 short inscriptions have the risk of being read incorrectly. This is true for a large number
932 of words in the Indus inscriptions. For example, tana can be read as naṭa, anu as unna,
933 etc. which may change the intended meaning. Table 14 shows examples where the
934 inscriptions could be read in the opposite direction with a different meaning. Scribes
935 eventually may have added the marks to err on the safe side, even if no obvious conflict
936 is known.
937 The most common sign, the  jar sign repeated sufficiently within inscriptions to
938 enable a directional marker of its own, by adding one to four small strokes inside the
939 jar as shown in the first section of Table 14. The strokes appear in non-ascending order.
940 The first jar always has the highest number of strokes and the next sign can either
941 decrease or maintain the number of strokes. Typically, the final jar sign will end up
942 with no strokes. This pattern holds for over two hundred inscriptions with just a handful
943 of exceptions as in the last section of Table 14. The improbability of these strokes to
944 have any kind of phonetic meaning is easily verified by testing the inscriptions against a
945 dictionary. These strokes are a directional marker similar to a fuel gauge, tracking the
946 remaining text portions from full to empty. The directional mark is placed on a nearby
947 sign if the starting sign is numeric or otherwise unsuitable for a directional indicator.
948 A directional marker may be used even when it’s the only jar sign in the inscription to
949 avoid accidental reading in the wrong direction as we see in the second section since the
950  jar sign is so often a terminal sign. When the leftmost sign has a directional marker
951 and the rightmost sign doesn’t, the seal may be preferentially read left to right as we
952 see in the last section of Table 14. This is most useful for multi-line boustrophedon
953 inscriptions but occasionally also occurs in single-line seals.
954 Signs other than jar signs can also have one to four strokes to indicate direction or
955 word boundary. Usually, these occur as the initial sign in the inscription and act as a
956 directional indicator and are listed in Table 10.

957 4.14. Stylistic variants

958 The different glyphs of the same allograph without directional function in Table 6 are
959 likely to be stylistic variations, based on the fact that different forms of a single allograph
960 rarely occur together in the same inscription. Typically when the inscription needs to
961 use the same allograph twice, the scribe seems to choose the same variant or one very
962 close to it. There are a few obvious exceptions, such as fish sign, where variants do
963 occur together.

964 4.15. Evolution

965 Signs that take up horizontal space undergo rotation      to fit in limited space.
966 Complex and time-consuming signs undergo abstraction      and simplification 
967  . Many variants of the same sign are simply cursive variants, as they are written
968 in a way to reduce the number of strokes on a softer medium. Indus script developed
969 several artifacts that can all be parsimoniously attributed to the lack of horizontal space
970 since inscribed objects were the size of coins. Eliding the final anusvāra (ṃ sound) and
971 double consonants were a matter of economy on small seals but are unnecessary on stone
972 and pillar inscriptions, given that the Brahmi anusvāra is a simple dot and doesn’t even

30
973 take up any space. This may be simply due to continued scribing tradition.
974 Examining the different glyphs attested for some signs shows evidence of the devel-
975 opment of cursive forms. For example, the signs that are based on straight lines show
976 curved lined variants such as  ,  ,  . This is unnecessarily complex in a carved
977 medium but natural in a soft medium using inks or brushes. The author of the inscrip-
978 tion seems to have written the text in cursive and the scribe must have carved it as is.
979 Minimizing strokes can sometimes work to the benefit of both carved and soft media,
980 such as  ,  , but some forms are clearly beneficial only on the non-carved media
981 such as  ,  . Ligatures also seem unnecessary on carved media but are natural
982 for ink and brush as in   .

983 4.16. Evolution to Brahmi

984 When we arrange Brahmi signs with their closest corresponding Indus script allograph
985 for the same sound value, we notice that every Brahmi sign seems to be a minor variant
986 or simplification of a pre-existing Indus sign. Indus script glyphs are closer to Brahmi
987 than Brahmi glyphs are to modern Devanagari script. Among the earliest Brahmi are
988 inscriptions from Keezhadi among which there is at least one sample on a potsherd that
989 reads 𑀫𑀱𑀭𑀼 𑀅. This is meaningless when read left-to-right but when read right-to-left,
990 reads अरुषम� aruṣam, which is a Rigvedic word for the red color of the god Agni, the Sun
991 or cows and horses[MW](Sivanantham and Seran, 2019). This is an early attestation of
992 direction change in Keezhadi. The gradual change from Indus script at the earliest levels,
993 followed by mixed inscriptions and finally only Brahmi on the latest levels may indicate
994 a gradual evolution from Indus script to Brahmi captured in Keezhadi stratigraphy.
995 Table 7 is an accurate snapshot of the Indus signs chosen to be standardized into
996 Brahmi using the observation that retroflexes and dentals are interchangeable as a group
997 as are non-labial nasal consonants.

998 4.17. Script fidelity

999 Bronze-age scripts generally do not have high fidelity to their transliteration. Linear
1000 B for example, has no consonant clusters. Sumerian is a polyvalent script that has
1001 multiple ways to read and write a word and different signs are used based on meaning.
1002 Even classical Arabic did not have dots (i’jām) to distinguish different phonemes on
1003 many rock inscriptions, despite having multiple forms (initial, media, terminal) for the
1004 same phoneme. By comparison, the Indus script has a fidelity similar to Tamil Brahmi
1005 with nearly the same sign reuse groups of consonants. Indus inscriptions are fairly
1006 readable even after 4-5 millennia with little effort.

1007 4.18. Religious continuity

1008 We see several linguistic and cultural features that continued post-Indus phase. The
1009 deities in Indus inscriptions are the same as Vedic. The Sun, Soma, Rudra, Indra, Agni,
1010 Ushas, Ashlesha and Bharani constellations, the horse ashva, Ardhanareeshwara (seals
1011 depicting one-breasted human) and pipal tree continue to be revered to this day. Select
1012 Rigvedic references related to long inscriptions are noted in Section 5.

31
 Table 15. Post-IVC attestation of script elements(Balasubramaniam, 2005; Bhatt, 1998; Hultzsch, 1925)

Indus feature Attestation Attested as Read as

merged dentals/ Avanti fish seal द� वट�ह devaṭāḥ द� वत�� devatāḥ

retroflex Mansehra दुवडश duvaḍaśa दुवदश duvadaśa

merged un/aspirated Girnar उस्ट�न usṭāna उस्थ�न usthāna

merged Kalsi द�श, द�ष dāśa, dāṣa द�स dāsa

sibilants Girnar वस� varsa वष� varṣa

no virāma Ajagara अगसतय agasataya अगस्तय agastya

elided doubled Qutub Iron pillar म�त्य�� mūrtyā म�त्त्य�� mūrttyā

consonants Girnar अ�ग agi अ��ग aggi

elided anusvara Girnar इद ida इद� idaṃ

अ� अन� aṃ an Qutub Iron pillar प्रन्श� pranśu प्र�श� praṃśu

flexibility Girnar अ� ञ aṃña अन्य anya

initial आ ā as अ a Girnar अर��पत� aropitaṃ आर��पत� āropitaṃ

1013 4.19. Vestiges of Indus linguistic forms

1014 Indus script characteristics are also attested into post-IVC inscriptions, coins and
1015 seals(Bhatt, 1998; Hultzsch, 1925; Solomon, 1998). These are shown in Table 15. The
1016 Iron pillar presently in the Qutub complex retains some of the archaic language features
1017 of Indus inscriptions(Singh, 2006). It uses a lone consonant to represent a doubled con-
1018 sonant (mūrtyā instead of mūrttyā) and the anusvara अ� aṃ is denoted अन� an (pranśu
1019 instead of praṃśu). Indus script provides evidence that the Qutub inscription was not
1020 an aberration, but rather that अ� aṃ was inscribed as अन� .
1021 Inscriptions with mixed Indus/Brahmi signs are now readable such as the one from
1022 Marungur(Subrahmanian, 2010). These are shown in Table 11. Indus-seal technology
1023 was exported to Mesopotamia and Dilmun in the mature phase(Laursen, 2010). Of the
1024 more than 121 seals unearthed, most do not have any inscription. At least 27 have Indus
1025 inscriptions shown in Table 12, of which some are readable as Sumerian, Akkadian or
1026 Old Akkadian(Association Assyrophile de France, 2006; Gelb, 1957; UPenn, 2006).

32
1027 5. Long inscriptions with their scriptural references

Seal-Id Inscription Translation

1 M-314    Honor the powerful Roarer

रव�मम� मन सक्षनर� जठलध�र रह O Sustainer, O Ocean(Shiva)
rava-amam mana sakṣa-naraṃ yield to the capable man
jaṭhala-dhāra raha
RV 5.45.10 hearing him, the waters receded

रव rava roarer[√रु + अच� ]; अमम� amam ams. the powerful[RV]; मन mana honor[√म्न� + ल�ट� 2s.];
सक्ष sakṣa capable[TS]; नर� naraṃ ams. man[TS]; जठल jaṭhala vms. Ocean[Sāy] =सम�द्र=Shiva[MBh];
ध�र dhāra vms. Sustainer[MBh 13.14.13]; रह raha yield, release[√रह� + ल�ट� 2s.];

2 M-23   Great Giver, this cup

द�ममन क�स� न� रन्धन is for us, O Shiva
dāma-mana kaṃsaṃ naḥ randhana
RV 2.37.2 his name is “The Giver”
RV 10.112.6 Filled is the beaker with the meath

द�मन� dāman giver[RV]; मन mana vns. great[=मह�न� mahān Vā.]; क�स� kaṃsaṃ ans. cup/jar[AV];
न� naḥ for us[RV dp. of अस्मद� asmad]; रन्धन randhana vms. destroying[Bhp];

3 M-234   Great Giver, the cup

द�ममन क�स� च आस नदन� is taken roaringly!
dāma-mana kaṃsaṃ ca āsa nadan
RV 10.136.7 Rudra roaring hath drunk from the cup

द�मन� dāman giver[RV]; मन mana vns. great[=मह�न� mahān Vā.]; क�स� kaṃsaṃ ans. cup/jar[AV];
आस āsa taken[√अस� + �लट� 1s./3s.]; नदन� nadan roaring[√नद� + शत��];

4 M-626   Excellent sacrificial cup

कम� म�ख� क�स� द�त नदन� O roaring purified one!
kam mākhaṃ kaṃsaṃ dāta nadan
RV 1.15.2 Drink from the purifier’s cup
RV 2.37.1 meath out of the sacrificer’s cup

कम� kam indc. good/well[TS]; म�ख� mākhaṃ ans. oblation[Hariv]; क�स� kaṃsaṃ ans. cup[AV];
द�त dāta vms. purified[Pāṇ]; नदन� nadan roaring[√नद� + शत��];

33
 Seal-Id Inscription Translation

5 H-1657  O Destroyer(Rudra), bring us

धक्क वह म�� सज�शवरम� the gift of food full of fatness
dhakka vaha māṃsaja-āśa-varam
RV 10.169.1 waters rich in life and fatness: to food that moves, O Rudra

धक्क dhakka vns. destroyer[√धक्क� + अच� ]; वह vaha carry[√वह� + ल�ट� 2s.]; म�� सज māṃsaja fat[Suśr];
आश āśa food[ŚBr]; वरम� varam ams. boon[RV];

6 M-355  O Roarer, That gift of

तद्ददनम� रव अममञ्जस्सर� powerful Soma juice
tat-dadanam rava amam-añjas-saraṃ
RV 9.80.2 powerful Soma, thou to whom the cows have-lowed

तत� tat ans. that[RV]; ददनम� dadanam ans. present[√दद� + ल्य�ट�]; रव rava vns. roarer[√रु + अच� ];
अमम� amamam ams. powerful[RV]; अञ्जस� añjas Soma[RV]; सरम� saram ams. liquid[VS];

7 M-38  O Roarer, Giver of plenty!

शतद�रव आम मह� म�ख� may I serve this great sacrifice
śatadā-rava āma mahaṃ mākhaṃ
RV 1.122.01 bringing sacrifice to bounteous Rudra

शतद� śatadā giver of a hundred[SV]; रव rava vns. roarer[√रु + अच� ]; आम āma serve[√अम� + ल�ट� 1s.];
मह� mahaṃ ams. great[RV]; म�ख� mākhaṃ ans. oblation[Hariv];

8 M-391  O burning roarer, bless us

द�हरव श� ज�र��ङ्कधर O lover, O Indra, O bearer
dāha-rava śaṃ jāra-aṅki-dhara
RV 1.115.2 Like a young man, Sun follows the Dawn
RV 3.45.4 Indra uses hooks to get fruit

द�ह dāha burning[MBh]; रव rava vns. roarer[√रु + अच� ]; श� śaṃ indc. bless[RV];
ज�र jāra lover of dawn[RV]; अ�ङ्कन� aṅkin Indra[RV 3.45.4]; धर dhara vms. bearer[MBh];

9 M-1954  O Shiva, soul of the day,

रव अहत्मन� जर�वर�न� bring us great praises
rava aha-tman jarā-varān
RV 1.26.9 may the eulogies of mortal men Belong to us

रव rava vns. roarer[√रु + अच� ]; अह aha day[RV]; त्मन� tman vms. soul[RV];
जर� jarā praise[RV]; वर�न� varān anp. best[RV];

34
 Seal-Id Inscription Translation

10 M-1688  O ancestor, strengthen indeed

वज ह�न आम� अश्वम�द�[न� ] the delights of possessing horses
vaja hāna ām aśva-modā[n]
RV 1.92.7 confer on us strength ... kine and horses

वज vaja strengthen[√वज� + ल�ट� 2s.]; ह�न hāna vms. departed[VS];

आम� ām indc. indeed[MaitrS]; अश्व aśva horse[RV]; म�द�[न� ] modān amp. joy[RV];

11 H-130  Grant us your power,

दद� तव वशम� अञ्जदह्र O Shining fire
dadā tava vaśam añja-dahra
RV 1.36.11 Agni, we extol: his powers shine out preeminent.

दद� dadā grant[RV 4.36.9]; तव tava your[RV]; वशम� vaśam ams. power[AV];
अञ्ज añja shining[√अञ्ज� + अच� ]; दह्र dahra vns. fire[Uṇvṛ];

12 M-359  Roaring Shiva, bring

तनभव वह अमम� वरम� a mighty blessing
tana-bhava vaha amam varam
RV 10.164.2 A happy boon do men elect, a mighty blessing they obtain

तन tana roarer[√तन� + अच� RV]; भव bhava vms. Rudra[BhP,AV]; वह vaha carry[√वह� + ल�ट� 2s.];
अमम� amam mighty[RV]; वरम� varam ams. choicest[MBh];

13 H-280  I enjoyed, O Shining one

हट ममद म�खवरम� the sacrificial gifts
haṭa mamada mākha-varam
RV 1.83.3 thy power brings blessing to the sacrificer pouring gifts.

हट haṭa vns. shining one[√हट� + अच� ]; ममद mamada enjoyed[√मद� + ल�ट� 1s.];
म�ख mākha oblation[Hariv]; वरम� varam ams. choicest[MBh];

14 M-369  O Subduing Spirit

दमनसत� मम म�खम� My sacrifice
damana-sat mama mākham
RV 10.99.6 subdued the demon who roared aloud

दमन damana subduing[MBh]; सत� sat vns. Universal Spirit[RV];

मम mama my[RV]; म�ख� mākhaṃ ams. sacrifice[Hariv];

35
 Seal-Id Inscription Translation

15 M-623  O Oblation bearer(Agni),

रस�लव�ह तमसहरण O darkness destroyer
rasāla-vāha tamasa-haraṇa
RV 1.12.2 Oblation-bearer, much beloved
RV 9.66.24 Destroying darkness black of hue

रस�ल rasāla oblation food of milk/curds[Īśvarasaṃhitā]; व�ह vāha vms. bearer[BhP];

तमस tamasa darkness[Uṇ., Sch]; हरण haraṇa vms. destroyer[Suśr];

16 M-10  O mighty bringer of wealth

स�त�वह�म अञ्जसम� तर bring Soma
sāta-āvaha-ama añjasam tara
RV 1.8.1 Indra, bring wealth
RV 2.14.8 to Indra bring ... the Soma.

स�त sāta wealth[RV]; आवह āvaha bringer[RV]; अम ama vms. mighty[RV];

अञ्जसम� añjasam ams. Soma[√अञ्जस� ]; तर tara bring[√त� + ल�ट� 2s.];

17 M-1169  [My] youngest from a hundred,

शत�त� अवमम� समञ्जन O shining one!
śatāt avamam sa-mañjana
RV 5.61.5 hundreds of sheep and steeds and kine

शत�त� śatāt hundred, abnp. शतम� śatam[RV]; अवमम� avamam ams. youngest[RV];
स sa indc. with[RV], मञ्जन mañjana vms. shine[√मञ्ज� + ल्य�ट�];

18 Ns-60  O unwavering shining bearer,

इद्ध�चलव�ह न�मनश� तर overcome the loss of rains
iddha-acala-vāha nāma-naśaṃ tara
RV 1.38.6 Let each (calamity), with drought, depart from us

इद्ध iddha shining; अचल acala immovable[RV,BG]; व�ह vāha vms. bearer[BhP];
न�म nāma from न�मन� nāman water[Naigh]; नश� naśaṃ ams. loss[MW]; तर tara overcome[√त� + ल�ट� 2s.];

19 M-714  O Roarer, ruler of

आरव श�स� अञ्जसभ� honoring council
ārava śāsaṃ añja-sabhaṃ
RV 1.114.2 Rudra, ruler of heroes

आरव ārava vns. Roarer[Pāṇ]; श�स� śāsaṃ ams. ruler[RV]; अञ्ज añja honoring[√अञ्ज� + अच� ];
सभ� sabhaṃ ans. council[RV];

36
 Seal-Id Inscription Translation

20 M-677  O Roarer,

आरव मम अञ्जस� म�न O my honored Soma
ārava mama añjas māna
RV 1.84.9 Soma juice prepared amid the many honours

आरव ārava vms. Roarer[Pāṇ]; मम mama my[RV];

अञ्जस� añjas vms. Soma mixture[RV]; म�न māna vms. honored[MBh];

21 M-577  Oh Desert Roamer(Indra)

आ धन्वचर नश�व�नन protect, O destroyer, O face
ā dhanvacara naśa ava ānana
RV 1.143.6 praise Him whose face is bright

आ ā indc. particle of reminiscence[Pāṇ.]; धन्वचर dhanvacara vms. desert Roamer[RV 5.36.1];

नश naśa vms. destroyer[√नश + अच� ]; अव ava protect[√अव� + ल�ट� 2s.]; आनन ānana vns. face[R];

22 M-665   O prayer-heeding fire

मतवच�द�हन त�म्र O red one
matavacaḥ-dāhana tāmra
RV 10.37.10 bless us with fervent heat and lustre

मतवचस� matavacas prayer-heeding[RV 1.46.5]; द�हन dāhana vms. fire[MBh];

त�म्र tāmra vms. red[MBh];

23 Lh-1  O fiery Roarer, help

द�वरव अम द�नर� the giver (of sacrifice)
dāva-rava ama dāna-raṃ
RV 1.114.4 We call Rudra for help, who fulfills our sacrifice

द�व dāva √दु fire[AV]; रव rava vns. roarer[√रु + अच� ];

अम ama help[√अम� ल�ट� 2s.]; द�न dāna giving[RV]; र� raṃ ams. effecting[Śiś];

24 K-10  New sacrifice oblation

नश्वरनवम�ख�
naśvara-nava-mākhaṃ
RV 10.89.3 I sing a new prayer
RV 10.89.17 may we partake of thy new favors

नश्वर naśvara perishable[Pur]; नव nava new[RV]; म�ख� mākhaṃ ans. oblation[Hariv];

37
 Seal-Id Inscription Translation

25 H-8  O Destroyer,

धक्क व अञ्जवरम� आच I seek a splendorous gift
dhakka va añja-varam āca
RV 1.92.13 bestow on us the wondrous gift

धक्क dhakka vns. destroyer[√धक्क� + अच� ]; व va indc. like; अञ्ज añja shining[√अञ्ज + अच� ];
वरम� varam ams. gift[RV]; आच āca I seek[√अच� + ल�ट� 1s.];

26 M-119  By the giver,

द� शतध�रवहमय[म� ] bearing a hundred streams
dā śata-dhāra-vahamaya[m]
RV 1.95.10 he makes stream ... and inundates with floods
RV 9.97.29 Sprung from the Gods, a hundred streams

द� dā ims. of द� dā giver[RV]; शत śata hundred; ध�र dhāra stream[RV 9.109.19];

वहमय[म� ] vahamaya[m] ans. bearing[Hcat];

27 H-103    O Kāmā-killer

हतमदन अञ्जन� आय[म� ] shining arrival
hata-madana añjan āya
RV 10.121.6 When over them the risen Sun is shining

हत hata Slayer[RV]; मदन madana vms. Kāmā[MBh]; अञ्जन� añjan shining[√अञ्ज� + शत��];
आय[म� ] āya[m] ams. arrival[RV]

28 M-494  Verily, the giftgiver sets flow

अन� आनष� स�त�द�
anā ānarṣa sātā-daḥ
RV 8.93.2 As from a mountain flow the water-brooks, thus flow his gifts

अन� anā indc. verily[RV]; आनष� ānarṣa set flow[√ऋष� + �लट� 3s.]; स�त� sātā gifts[RV];
द� daḥ nms. giver[MBh];

29 H-268  O extraordinary one

दस्म वह अञ्जस्भ�� flow-forth the shining Soma
dasma vaha añjas-bhāṃ
RV 8.96.10 brightly-shining, mingled with the milk, the draughts of Soma

दस्म dasma vms. extraordinary[RV]; वह vaha bring[√वह� + ल�ट� 2s.]; अञ्जस� añjas Soma[RV];
भ�� bhāṃ ans. shine[RV];

38
 Seal-Id Inscription Translation

30 M-396  O unequalled one,

असम�मत� आत्मतस� serve from the self
asama amatu ātmatas
RV 4.12.3 the self-reliant God,
RV 2.3.1 Let Agni serve the gods

असम asama vms. unequalled[RV]; अमत� amatu serve[√अम� + ल�ट� 3s.];

आत्मतस� ātmatas abns. from the self[आत्मन� + त�सल� RV];

31 M-1676  O Savior

समव वम तम�मरम� emit darkness removing (light)
samava vama tamaḥ-maram
RV 1.62.5 with the Dawn, Sun, rays, dispelled the darkness
समव samava vms. helper[समव� + अच� RV]; वम vama[√वम� + ल�ट� 2s.];
तमस� tamas darkness[RV]; मरम� maram ams. death[RV];

32 K-15   O my healer(Shiva), with roofs

मम रसद सछद��न अक��र may I raise, O Sun
mama rasada sa-chadāni arka-āra
RV 4.5.1 he uplifted as a pillar bears the roof

मम mama my[RV]; रसद rsada vms. healer[MBh]; स sa indc. with[RV];

छद��न chadāni amp. cover[BhP]; अक� arka vms. Sun[RV];आर āra raise[√ऋ + ल�ट� 1s. RV];

33 M-1690  O Yama’s killer(Shiva),

यमहन� अ� श� शममन� My portion, O calm one
yama-han aṃśaṃ śama-man
RV 3.62.11 Our portion of prosperity

यम yama Lord of death[RV]; हन� han vms. killer[RV]; अ� श� aṃśa ams. portion[RV 2.19.5];
शम śama calm[MBh]; -मन� -man vms. containing[R];

34 Ns-9   Verily, O immovable one

अथ अचलत्व म�न O honored one
atha acala-tva māna
RV 8.14.9 Firm and immovable from their place

अथ atha indc. now/hereby[RV]; अचल acala immovable[Bhag]; त्व tva vms. one[RV];
म�न māna vms. honored[MBh]

39
 Seal-Id Inscription Translation

35 H-12  O Roarer, a powerful

रव अमम� महम�ख� great sacrifice (to you)
rava amam maha-mākhaṃ
RV 1.75.5 our mighty sacrifice

रव rava vns. Roarer[√रु + अच� ]; अमम� amam ams. powerful[RV]; मह maha great[RV];
म�ख� mākhaṃ ans. oblation[Harv];

36 M-495  Verily, he set flow the gifts

अन� आनष� स�त�न�
anā ānarṣa sātān
RV 8.93.2 As from a mountain flow the water-brooks, thus flow his gifts

अन� anā indc. verily[RV]; आनष� ānarṣa set flow[√ऋष� + �लट� 3s.];
स�त�न� sātān amp. gifts[RV];

37 M-900  O Roarer, my lasting wish

रव सहत� मम वरम�
rava sahat mama varam
RV 6.45.21 satisfy our wish with power and wealth

रव rava vns. Roarer[√रु + अच� ]; सहत� sahat ans. lasting[RV]; मम mama my[RV];
वरम� varam ams. wish[RV];

38 M-52  O subduer,

दमवह आ मथमह�न� O great Churner
dama-vaha ā matha-mahān
RV 10.24.4 ye Twain churned the united worlds apart

दम dama subdue[RV]; वह vaha vms. bearing[RV]; आ ā as[RV], मथ matha churner[√मथ + अच� ];

मह�न� mahān vms. महत� great[RV];

39 C-24    O Roaring destroyer, salutations

रसतस नमन� स� ध�र O possessor of all things
rasa-tasa naman saṃ-dhāra
RV 4.8.1 Your envoy who possesses all
रस rasa roarer[√रस� + अच� ŚBr]; तस tasa vms. destroyer[√तस� + अच� Dhātup];
नमन� naman saluting[√नम� + शत��]; स� saṃ indc. all[RV]; ध�र dhāra vns. possessing[BG];

40
 Seal-Id Inscription Translation

40 H-20  I am capable of salvation

अहम� सह� म�क्षम�
aham sahaṃ mokṣam
RV 10.52.5 I shall give you freedom

अहम� aham ns. I[RV]; सह� sahaṃ ams. capable[Kālid]; म�क्षम� mokṣam ams. salvation[MBh];

41 Krs-1  O Ocean, (protect)

नम� जठल श�य[म� ] the sleeping Nami
namīm jaṭhala śāya[m]
RV 6.20.6 He guarded Nami, Sapya’s son, in slumber

नम� namīṃ ams. Nami, son of Sapya[RV 6.20.6]; जठल jaṭhala vms. Ocean;
श�य[म� ] śāya[m] ans. sleeping[RV];

42 M-49  Indeed, pulverizing

अह अचलवशम� मषम� mountain at will
aha acala-vaśam maṣam
RV 1.55.3 great manly power, so as to bend ... that famed mountain down

अह aha indc. Indeed[RV]; अचल acala mountain[GārgīS];

वशम� vaśam ams. at will[RV]; मषम� maṣam indc. pulverized[ŚāṅkhGṛ];

43 M-1052  I served small offerings

आम व�हस्मन�क�
āma vāhas-manāk
RV 10.91.8 sacrificial offerings great and small alike.

आम āma I served[√अम� + �लट� 1s.]; व�हस� vāhas offering[RV];

मन�क� manāk indc. small[Kāv];

44 M-634  bestowed fit small chariot

ऋतरथर�तमन�क�
ṛta-ratha-rāta-manāk
RV 8.58.15 boy exceeding small, mounted his newly-fashioned car.

ऋत ṛta fit[RV]; रथ ratha chariot[RV];

र�त rāta bestowed[RV]; मन�क� manāk indc. small[Kāv];

41
 Seal-Id Inscription Translation

45 H-129  do invite, O powerful

भसरव�म आम� त्र shining Roarer
bhasa-rava-ama āmantra
RV 6.50.4 This day invited the Sons of Rudra

भस bhasa shining[√भस + अच� ]; रव rava Roarer[√रु + अच� ]; अम ama vms. powerful[RV];

आमन्त्र āmantra invite[आ + √मन्त्र + ल�ट� 2s.];

46 M-1834   Giving and restoring

दद�न�दम� जननल Speech maker (Agni)
dada-anudam jana-nala
RV 10.137.1 restore to life again the man who has committed sin
RV 3.1.19 make our share glorious and adorned with fine speech

दद dada giving[RV]; अन�द� anudaṃ ans. restoring[RV]; जन jana creator[√जन� + अच� ];

नल nala speaker[√नल� + अच� ];

47 Nd-1  O red bodied one

शमशर�ङ्गतम्र with calmed arrows
śama-śara-aṅga-tamra
Sri Rudram/KYV TS 4.1 Your calm arrow ... O red one

शम śama calm[MBh]; शर śara arrow[RV,MBh]; अङ्ग aṅga body[Uṇ]; तम्र tamra vms. red/dark[RV];
[Sri Rudram: 1.4 य�त इष�� �शवतम�, 1.11 यस्त�म्र�]

48 M-1892   Suitable bountiful rains

अरवष��रर�ण�न�
ara-varṣā-rarāṇān
RV 2.27.15 shed in abundance, The rain of heaven
अरम� aram indc. suitable[RV]; वष�� varṣā rain[TS]; रर�ण�न� rarāṇān anp. bountiful[RV];

49 H-270  I salute the praised destroyer

ननम शस्त� जष�
nanama śastaṃ jaṣaṃ
RV 8.86.10 Most eminent for power, destroyer in the conflict
RV 10.170.3 he spread wide unfailing victory and strength

ननम nanama I salute[√नम� + ल�ट� 1s.]; शस्त� śastaṃ ams. praised[MBh];

जष� jaṣaṃ ams. destroyer[√जष� + अच� ];

42
 Seal-Id Inscription Translation

50 H-10  I steadied the strong horse

बबद अश्ववलर[म� ]
babada aśva-vala-ra[m]
RV 1.38.12 steady your horses

बबद babada steadied[√बद� + �लट� 1s.]; अश्व aśva horse[RV];

वल=बल vala=bala strength[RV]; रम� ram ams. possessing[Naiṣ];

51 M-1123  Given speed

रर ररय[म� ]
rara ra-raya[m]
RV 2.31.2 Gods of one, accord speed on our car

रर rara given[√र� + �लट� 2p.]; र ra possessing[RV];

रय[म� ] raya[m] ams. speed[Pur];

43
1028 6. Sign Variants and Conjuncts

1029 Variants are stylistic and abstracted evolution of symbols. Conjunct signs combine two
1030 or more signs and are read as adjacent signs: firstly in the direction of writing, secondly
1031 top to bottom and lastly the base character followed by the ligatured character.

1032 (1) Ligatured symbols that are simply touching but otherwise written in normal size
1033 and position are read as normal. For example,  is simply read as .
1034 (2) Symbols arranged vertically are read from top to bottom.  is read as  �र as
1035 opposed to  which is read as  ईर
1036 (3) The base character, i.e., the large character is read first and the embedded char-
1037 acter is read subsequent.  becomes  रस� .

Sign Variant Inscription Sanskrit

अ M-875→  आम्र
 
a āmra
1
mango[MBh]

अ M-403  अन्तर

 
a antara
2
nearby[RV]

अ H-777  अचर

 
a acara
3
immovable[RV]

अ H-829→  अवर

 
a avara
4
youngest[RV]

अ M-210  अरण

 
a araṇa
5
distant[RV]

अ M-123  मन� �

 
a manaiḥ
6
by the great one[Vā]

आ M-1151  आम� नमन�

 
ā ām naman
7
verily[MaitrS] worshipping[√नम� ]

अक M-1764  अस्म�क[म� ]

 
aka asmāka[m]
8
ours[RV]

आक� H-2104  आक�शम�

 
ākā ākāśam
9
sky[ŚBr]

अद� Ad-6 /\ तद्धनद�म

/ \ 
adā tat-dhana-dāma
10
giver of that wealth[RV,RV,RV]

अन� M-89  अमह�न�न

 
an amaha-anūna
11
O complete[AV] giver[RV]

44
 Sign Variant Inscription Sanskrit
अन� H-228  आन�य�
 
an fisherman [nms. आन��यन� ] ānāyī
12

अन L-218  अनवस

 
ana anavasa
13
unstoppable[RV]

अन� M-812  अन�म�

 
anā anām
14
indeed, indeed[RV,MaitrS]

अन्न C-68  अन्नद अञ्ज

 
anna annada añja
15
speak[अञ्ज� ], O foodgiver[MS]

अन्तर M-639  अन्तर�श्वत्थ नमन�

 
antara antara-aśvattha naman
16
saluting Ashwatta[AV] spirit[RV]

आप H-413  आप
 
āpa āpa
17
vedic deity(Vasus)[MBh]

आभ� M-742  अभ�म� सर�[म� ]

 
ābhā ābhām sarā[m]
18
Splendorous[RV] streams[AV]

अम M-18  जनम�

 
ama janam
19
people[RV]

अम M-896  अम�श�य

 
ama ama-āśāya
20
sleeping[AV] at home[RV]

अम D-51582  आभ�मह

 
ama ābhāmaha
21
O shining[RV] day[RV]

अम M-969  आमय
 
ama serve[√अम� + ल�ट� 2s.] āmaya
22

आम H-449  अमरणम� वर�

 
āma a-maraṇam varaṃ
23
gift[RV] of immortality[MBh]

अम� RGR-7230  आम�न

 
āmā āmāna
24
to honor[Mn]

अर M-713  आरवदहनमन

25   ara great[Vā] burning[√दह�] Roarer[√रु] ārava-dahana-mana

अर K-45→  अण�पय

 
ara arṇapaya
26
power[RV] of flood[RV]

45
 Sign Variant Inscription Sanskrit
आरव M-1105  आरव
 
ārava ārava
27
Roarer[√रु]

आल M-367    आलवहनमन

 
āla āla-vaha-namana
28
salute[RV] poison[Suśr] remover[RV]

अश� H-1513  अश्न

  aś aśna
29
voracious[RV]

अश� M-821  अशम�

  aś aśam
30
eating[ŚBr]

अश� H-1841  मसद�

  aś masadaṃ
31
to the tansformer[√मस� ]

अश� H-1994  भ�� श

 
aś bhāṃśa
32
ray[RV] of light[MBh]

अश� H-1049  अश्वचर

 
aś aśva-cara
33
O mover[MBh] on horses[RV]

अश� M-243  दत्त�श्वद

 
aś datta-aśva-da
34
given[RV] by horsegiver[Mn]

अस� M-954  मम अनस�

 
as mama anas
35
my[RV] cart[RV]

अस M-1724  आशभ�ज्य

  asa āśabhājya
36
food[ŚBr] to share[BhP]

आस� K-17  आश�

 
ās āśaṃ
37
food[ŚBr]

अ�श H-679  श�श[न� ]

  aśi śaśi[n]
38
moon[MBh]

अश M-219  अश्म[न� ]
 
aśa aśma[n]
39
sky[RV]

आश M-2088  आश
 
āśa āśa
40
food[ŚBr]

आश M-1887  म�म� आश�

 
āśa mām āśaṃ
41
obtaining me[RV]

46
 Sign Variant Inscription Sanskrit
आश� H-156  आश��
 
āśā āśāṃ
42
desires[AV]

अश्व M-1106  अन� अश्वद

 
aśva anā aśvada
43
indeed[RV] horse[RV] giver[RV]

ईर M-1170  ईर
 
īra īra
44
rising [ईर� :RV]

ई�र M-151  ई�रणम�

 
īri īriṇam
45
desert[MBh]

ईशन� M-1548  सह ईशन�

 
īśan saha īśan
46
Vanquish[RV], O Shiva[MBh]

उ M-20  उदसव�पर

 
u uda-sarva-para
47
Beyond[RV] all[RV] waters[RV]

क H-66  क�मनन�

 
ka kāmanan
48
misremembering[Pāṇ,√म्न�]

क H-513  क्षण

 
ka kṣaṇa
49
moment[Śak]

क M-68↔   दमनह�नक�म

 
ka damana-hāna-kāma
50
subduing[MBh,Gaut] of Kāma

कर M-57  खरव�हमह नमन

 
kara khara-vāha-maha namana
51
saluting cloud[R] bringer[MBh]

च H-214  �छ�न्न[न� ]

 
ca chinni[n]
52
destroyer[�छन्न:AV]

चर C-13  चम�समम�

  cara carma-samam
53
shield[MBh] like[RV]

ज M-1848  जष
 
ja jaṣa
54
O destroyer[√जष� Dhātup]

जर M-898  जरम�
 
jara jaram
55
praise[√ज� RV]

जस� Ns-4  यजस�

 
jas yajas
56
sacrifice[RV]

47
 Sign Variant Inscription Sanskrit
�ज M-409  स�ञ्ज
 
ji sañji
57
complete victory[RV]

/ त K-11 /
 आतटद�
 
ta ā-taṭa-daṃ
58
one who crosses[MBh]

त M-1082  तन्व आमत�

 
ta tanva āmat
59
Tanva has served[√अम� ]

तर M-1792  अन्तर� मर�

 
tara antaraṃ maraṃ
60
near[RV] death[RV]

तर H-1192  तमत्र��ण

  tara tama-trāṇi
61
dark[RV] and protected[MBh]

त�स M-1721  रवत�सम�

   
tāsa Roarer’s[√रु] destruction[√तस� घञ� ] rava-tāsam
62

त्र M-360  अमत्रद�न

  tra amatra-dāna
63
powerful[RV] gift[RV]

द H-7  दद एकपद�

 
da dada ekapadaḥ
64
from Vishnu[MBh], O Giver[Paṇ]

द M-1793  दम� नमन�

 
da damaḥ naman
65
house[RV] praying[ŚBr]

द H-792  दय�

 
d kind one[ŚBr दय�+इ�न] dayī
66

ध M-73  धन�

 
d dhanī
67
wealthy one[RV √धन+इ�न]

धक्क M-261  धक्क अतन

 
dhakka Destroyer[√धक्क�] passing[Nir] dhakka atana
68

द� H-1662   द�व तत�न�

 
dā dāva tatān
69
O fire[MBh], to my fathers[RV]

द�द Umma  द�द�ण्वचर

 
dāda dāda-aṇvara-cara
70
gift[MBh] of Soma[RV,MBh]

ध� M-605  ध�
 
dhī dhī
71
devotion[RV]

48
 Sign Variant Inscription Sanskrit
न H-74  सद�न्व
 
na sadānva
72
roaring[RV]

न H-295  सन
 
na sana
73
ancient[RV]

न M-1277  नश
 
na naśa
74
destroyer[√नश� ]

न M-747  न न�श�

 
na na nāśaṃ
75
no loss[MBh]

न M-986  नर�

 
na naraṃ
76
man[TS]

नम� H-1869  नम�न्य

 
namā nama-anya
77
inexhaustible[AV] pasture[RV]

न�शन M-975  नशन�नलसक्ष

 
naśana naśana-anala-sakṣa
78
loss[BhP] by intense fire[Subh]

प H-1024  मव पण
 
pa mava paṇa
79
bind[√मव� ] the wager[MBh]

पद M-777 
 असम�मनपद

 
pada asama-āmana-pada
80
unequalled[RV] friendly[TS] sign[MBh]

पर M-83  प्रसर

 
pra prasara
81
stream[BhP]

बत Frm-1329  भटन�
 
bata employing[√भट� +शत��] bhaṭan
82

बद्ध K-53  बद्ध

 
baddha baddha
83
bound[RV]

भ M-331  आभस�

 
bha ābhasa
84
splendor[आभ�स ābhāsa R]

म H-2125  म�न

 
ma māna
85
honor[MBh]

49
 Sign Variant Inscription Sanskrit
म M-523  समर
 
ma samara
86
confluence[RV]

म M-1367  मर
 
ma mara
87
death[AV]

मद C-8  मद� मह�

 
mada madaṃ mahaṃ
88
abundant[RV] joy[RV]

मर H-142  कम� तम्र�त�न

 
mara kam tamra-ātān
89
pleasant[TS] spreading[RV] darkness[RV]

म� M-649  म�र
 
mī mīra
90
ocean[Uṇ]

य H-951  वनयव

 
ya vana-yava
91
wild[RV] grain[RV]

य M-831  यत
 
ya yata
92
controlled[RV]

य H-455  हयव�न� जन

 
ya hayavān jana
93
O horse-owning[RV] people[RV]

य C-39  दध्य

 
ya dadhya
94
meditate[√ध्य� ]

य अय M-288  य�यनम� शरण[म� ]

 
ya aya yāyanam śaraṇam
95
protecting[RV] journey[RV]

ऋभ� M-2012  ऋभ्वप�न

 
ṛbh ṛbhva-pāna
96
Indra’s[RV] drink[RV]

र K-18  तद� वरल

  ra tad varala
97
that[AV] swan[Mṛga]

र H-212  र�य

 
ra rayi
98
wealth[RV]

र H-510  रमवरम�

 
ra rama-varam
99
delightful boon[RV]

50
 Sign Variant Inscription Sanskrit
र L-48   ररख अ� श नमन�
 
ra rakha aṃśa naman
100
I moved[√रख� ] praying[√नम� ] O Aditya[RV 2.1.4]

र H-299  रण्व

 
ra raṇva
101
delightful[RV]

र M-1424→  रणभ�द

 
ra raṇa-bhāda
102
O pleasing[RV] lightgiver[MBh]

र H-2214→  रसमत�

 
ra rasamat
103
juicy[RV]

र Salut  व्यवचर

 
ra vyavacara
104
move[RV] away[√व्यय� ]

र H-1988  इमम� रजम�

 
ra imam rajam
105
this[RV] color[Sarvad]

र H-172  प��णधर

 
ra pāṇidhara
106
O hand holder[RV,Kāv]

रक K-15/2  अक��र

 
raka arka-ara
107
reached[√ऋ] the Sun[RV]

रक M-120   मदनब�ण रख

 
raka madana-bāṇa rakha
108
go, O cupid’s[MBh] arrow[RV]

रक्त M-1759  रक्तवधहन�

  rakta rakta-vadhahan
109
O red one[MBh], who kills[RV]

रण M-1771  रण� भ��

 
raṇa raṇaṃ bhāṃ
110
delightful[RV] lustre[MBh]

रर H-405  रर वहन�मन�

 
rara rara vahanāman
111
give, O sign[RV] of bearer[RV]

रर M-1154  प�शमन रर

 
rara pāśamana rara
112
Give[√र�], O great[Vā] noose[RV]

रज K-122  रर जट
 
raja rara jaṭa
113
granted[√र�] O Shiva[MBh]

रस� H-598  रहम� अवमतम�

  ras raham avamatam
114
separator[√रह�] of despised[Mn]

51
 Sign Variant Inscription Sanskrit
र� M-84  धन अमर�न�
115   rā O producer![Dhatup] the immortals[ŚBr] dhana amarān

�र Desalpur  �रशवधम�

 
ri riśavadham
116
destroyer[AV] killer[RV]

�र M-918  रव र�र

 
ri rava rīra
117
O Roarer[√रु], O Shiva[Cat]

र� H-1033  हर�वन�म

 
rī hari-iva-nāma
118
Golden[RV,RV] appearance[RV]

लल M-751A  ललक
 
lala lalaka
119
little delight[√लल� ]

व MS-5062  अवभ�स�

 
va avabhās
120
shine forth[MBh]

श M-207  शद
 
śa śada
121
Destroyer[√शद� MBh]

स Louvre  आसज अयतस�

 
sa āsaja-ayatas
122
unrestrained[RV] speed[RV]

सक्ष M-1342  सभ�सक्ष�

 
sakṣa sabhā-sakṣaṃ
123
powerful[TS] council[RV]

स�त H-325  स�त�

 
sāta sātaṃ
124
gains[RV]

सन M-800  चषन� श नमन�

 
sana caṣan-aṃśa naman
125
saluting[MārkP] destroyer[√चष� ,RV]

सर D-9824  शव��म मञ्ज

 
sara śarva-ama mañja
126
Shine[√मञ्ज� ] mighty[RV] Shiva[AV]

ह� M-203  मह� अनवर�

 
hā mahā anvaraṃ
127
great[RV] excellence[RV]

52
1038 7. Very Short Inscriptions

Seal-Id Inscription Translation Seal-Id Inscription Translation

1 Bhirrana  अम� to fix[TS] 2 M-1203  अम� to fix[TS]
3 H-367  अम� to fix[TS] 4 L-106  अस shine![√अस� ]
5 M-842  अ�स sword[RV] 6 M-1198  अ (voc.)[T]
7 M-1905  अ (voc.)[T] 8 H-481  आ (voc.)[Pāṇ]
9 C-52  अक्क O mother[Pāṇ] 10 M-1105  आरव O Roarer[√रु]
11 K-464  अश master[RV] 12 M-597  ईश O Shiva[MBh]
13 H-1113  इ (interj)[Pāṇ] 14 M-949  क joy[Nir]
15 H-1113  क joy[Nir] 16 H-1833  च also[Pāṇ]
17 M-607  च also[Pāṇ] 18 M-262  अच beg[√अच� ]
19 M-331  ज born[Mn] 20 M-85-5 / त[म� ] him[RV]
21 D-50318  त[म� ] him[RV] 22 L-66  त[म� ] him[RV]
23 Blk-5  त[म� ] him[RV] 24 H-94  त�[न� ] them[MS]
25 M-996  द giving[MS] 26 M-326  द giving[MS]
27 M-470  द giving[MS] 28 H-1514  द giving[MS]
29 H-1011  द giving[MS] 30 K-476  द�द gift[MBh]
31 M-605  ध� prayer[RV] 32 M-326  न praised[TS]
33 M-1118  नमन bow[MārkP] 34 M-1084  न�मन� sign[RV]
35 D-48576  नवन praising[Nalod] 36 K-53  बद्ध bound[RV]
37 M-273  भ O star[Sūryas] 38 K-67  भ O star[Sūryas]
39 L-105  बभ shining[√भ�] 40 M-1465  प protection
41 M-410  म� to me[म�म� ] 42 M-1898  म� to me[म�म� ]
43 L-54  म� to me[म�म� ] 44 M-516  म� to me[म�म� ]
45 M-1162  य him[यद�] 46 Lakhanjo  य him[यद�]
47 K-446  य him[यद�] 48 C-94  य him[यद�]
49 M-1563  र� giving[Śiś] 50 M-604  र� giving[Śiś]
51 B-12  र giving[Śiś] 52 M-1641  र giving[Śiś]
53 M-1205  र giving[Śiś] 54 M-1233  र giving[Śiś]
55 M-593  र giving[Śiś] 56 M-599  र giving[Śiś]
57 K-462 ↑ ह verily[Ved] 58 D-17490  ह verily[Ved]
59 M-1642  व similar[MBh] 60 K-458  ह verily[Ved]
61 L-66  ह verily[Ved] 62 B-10  ह verily[Ved]
63 Bhirrana  ह verily[Ved] 64 H-1016  ह verily[Ved]

53
1039 8. Derivation

1040 8.1.  · अन� · an from अ� श� aṃśu soma drink

Seal-Id Inscription Sanskrit Translation

H-764 B  अनन� ananaṃ the living[Nir]

Dmd-1  अन ana breath[ŚBr]
(Many) * *अ� *aṃ (terminal अन�स्व�र)

 = अन� अ� · an am (1)

1041 8.2.  · अ · a from आज�न ājani stick

Seal-Id Inscription Sanskrit Translation

Harappa  आननम� ānanaṃ face[R]

H-1550  अन� anā indeed[RV]
H-1919  आम� ām verily[MaitrS]

 = अ · a from 1 (2)

1042 8.3.  · द · d from धन्वन� dhanvan bow

Seal-Id Inscription Sanskrit Translation

C-80  अन्ध andha O Soma[RV]

H-1919  धन dhana wealth, prize[RV]
M-1637  ध� dhā to place[RV]

 = द · d from 1, 2 (3)

54
1043 8.4.  · अ · a from आय� āyu man

Seal-Id Inscription Sanskrit Translation

M-191  ध� dhā to place[RV, कण� कर� ध�]

H-346  आम� ām verily[MaitrS]

 = अ · a from 1, 3 (4)

1044 8.5.  · इ · i from इष�क iṣīkā stalk of grass

Seal-Id Inscription Sanskrit Translation

M-87A  अ�ध adhi above[RV]

H-2244B  इद� id (affirmation)[Ved]

 = इ · i from 3, 4 (5)

1045 8.6.  · द · d from धन्वन� dhanvan variant of bow 

Seal-Id Inscription Sanskrit Translation

C-15a  अध�न adhīna belonging to[R]

 = द · d from 1, 2, 5 (6)

1046 8.7.  · त · t from त�ड� ल taḍula variant of fighter 

Seal-Id Inscription Sanskrit Translation

B-8  तन tana offspring[AV]

H-150  ददद�न्त dada-dānta Given[RV 1.39.9] by
pacified one[√दम� + क्त]

 = त · t from 1, 6 (7)

55
1047 8.8.  · श · s from श�क्र śukra seed

Seal-Id Inscription Sanskrit Translation

M-916A  शश� śaśī the Moon[ŚvetUp]

M-482A  श�न śani Saturn[R]

 = श · s from 1, 5 (8)

1048 8.9.  · न · n from न�ल nāla stalk

Seal-Id Inscription Sanskrit Translation

M-992a  अशन aśana reaching across[Nir]

M-812a  अन�न� anān breathing[ŚBr]

 = न · n from 1, 2, 8 (9)

1049 8.10.  · ई · ī from इइ ī long i

Seal-Id Inscription Sanskrit Translation

M-1824  अनन� ananī the living[anana + इ�न]

 = ई · i� from 1, 5 (10)

1050 8.11.  · ज · j from झर jhara waterfall; cascade

Seal-Id Inscription Sanskrit Translation

H-246B  जनध� janadhā nourishing creatures [TBr]

 = ज · j from 1, 3, 4 (11)

56
1051 8.12.  · र · r from रथ ratha chariot

Seal-Id Inscription Sanskrit Translation

H-842  ऋण� ṛṇī debtor[BG]

H-923  जर jara aging[RV]
H-305B  अरज araja dustless (pure)[R]

 = र · r from 1, 2, 5, 11 (12)

1052 8.13.  · र · r from रथव� ratharvī multi/split snake

Seal-Id Inscription Sanskrit Translation

H-1745A  रज raja pollen[Prasaṅgābh]

H-585A  जर jara aging[RV]
M-1170A  ईर īra mover[√ईर� + अच� RV]

 = र · r from 5, 11 (13)

1053 8.14.  · च · c from चत�र� catur four

Seal-Id Inscription Sanskrit Translation

D-33544  च�र cari (a name)[Pravar]

M-749A  चर cara move[MBh]
H-215B  चण caṇa chickpea[MBh]

 = च · c from 1, 5, 13 (14)

57
1054 8.15.  · अ · a from अजशृङ्ग� ajaśṛṅgī goat’s horn

Seal-Id Inscription Sanskrit Translation

L-20a  अचर acara immovable[RV]

H-70a  अजर ajara undecaying[RV]
H-2173  अर�ण araṇi kindling wood[RV]

 = अ · a from 1, 5, 11, 12, 13, 14 (15)

1055 8.16.  · म · m from मत्स्य matsya fish

Seal-Id Inscription Sanskrit Translation

M-238a  मसन masana transformation[√मस� + ल्य�ट�]

M-1344A  मरण maraṇa death[MBh]
H-1192A  चमन� caman eating[√चम� + शत��]

 = म · m from 1, 8, 13, 14 (16)

1056 8.17.  · व · v from �व vi two

Seal-Id Inscription Sanskrit Translation

H-292A  रव rava O Roarer[√रु + अच� ]

M-330  अन� अनव� anā anarva Indeed, O unstoppable one[RV,RV]

 = व · v from 1, 12 (17)

1057 8.18.  · र · r from रथ ratha variant of ratha 

Seal-Id Inscription Sanskrit Translation

M-1233  वण�ज varṇaja born of class[VarBṛS]

 = र · r from 1, 11, 17 (18)

58
1058 8.19.  · र · r from रथ��रन� rathārin chariot wheel

Seal-Id Inscription Sanskrit Translation

M-1119a  अनव� श� anarva-śaṃ Bless, O unstoppable one[RV,RV]

H-2018  रज raja passion[Uṇ.]

 = र · r from 1, 8, 11, 12, 17 (19)

1059 8.20.  · ब · b from भक्षपत्त्र� bhakṣapattrī betel leaf

Seal-Id Inscription Sanskrit Translation

M-311A  भणव� bhaṇavī Roarer[√भण� + अच� + व� Pāṇ,RV]

 = ब · b from 5, 9, 17 (20)

1060 8.21.  · य · y from यव�� �� yavaśreṣṭhi grain merchant

Seal-Id Inscription Sanskrit Translation

H-81A  जय jaya victory[AV]

M-209A  म�न्य mānya respected[MBh]

 = य · y from 1, 11, 16 (21)

1061 8.22.  · अ · a from आय� āyu variant of 

Seal-Id Inscription Sanskrit Translation

L-38  म�य māya illusion[MBh]

M-1654  आन āna nose[RV]

 = अ · a from 1, 2, 16, 21 (22)

59
1062 8.23.  · न · n from न�ल�क� nālīka arrow

Seal-Id Inscription Sanskrit Translation

M-1206  चमन camana eating[Jaim]

M-403  अन्तर antara internal[RV]

 = न · n from 4, 7, 13, 14, 16 (23)

1063 8.24.  · अस� · as from अष्टप�द aṣṭapāda eight legged; spider

Seal-Id Inscription Sanskrit Translation

H-1513  अश्न aśna voracious[RV]

B-1A  ररस्नवमन ra-rasna-vamana Emitting[Kālid,Śiś] object[Uṇ]

 = अस� · as from 1, 4, 9, 16, 17, 19 (24)

1064 8.25.  · क · k from क�तम� kṛtam dice

Seal-Id Inscription Sanskrit Translation

H-300B  रणक raṇaka N. of King[BhP]

M-176  कम� kam well[ŚBr]

 = क · k from 1, 9, 12 (25)

1065 8.26.  · अ · a from अङ्क aṅka curve

Seal-Id Inscription Sanskrit Translation

H-1520  आजक ājaka goats[Pāṇ]

M-1658  आभ ābha beauty[MBh]
Nuhato  श�ण śāṇa flaxen[ŚBr]

 = अ · a from 1, 8, 11, 20, 25 (26)

60
1066 8.27.  · अम · ama from अङ्क + मत्स्य ama conjunct curve + fish

Seal-Id Inscription Sanskrit Translation

M-686a  स�मन� sāman wealth[RV]

H-1085  ज�म� jāmā daughter[MBh]
Allahdino  अम� amaṃ power[RV]

 = अम · ama from 1, 8, 11, 23 (27)

1067 8.28.  · न · n from न�ल nāla reed

Seal-Id Inscription Sanskrit Translation

H-59A  अश्न� aśnaṃ a stone[RV]

M-1118A  नमन namana salute[MārkP]

 = न · n from 1, 16, 24 (28)

1068 8.29.  · अन� · n from अ� श� aṃśu lamp

Seal-Id Inscription Sanskrit Translation

H-550  आञ्जन āñjana fat[RV]

M-928A  आनम� ānam subdue[RV]
Harappa  अञ्च añca curling[RV]

 = न · n from 2, 11, 14, 23, 27 (29)

1069 8.30.  · त · t from त�� tardū variant of wooden ladles 

Seal-Id Inscription Sanskrit Translation

H-786  त्मन� tman self[RV]

L-145  अश्वतर aśvatara mule[AV]

 = त · t from 13, 16, 17, 23, 24 (30)

61
1070 8.31.  · स · s from स�प�न sopāna ladder

Seal-Id Inscription Sanskrit Translation

H-1492  शश śaśa rabbit[RV]

H-1533  अ� श aṃśa stake[RV]
H-2585  श�न śani Saturn[R]

 = स · s from 1, 5 (31)

1071 8.32.  · अ · a from अङ्क aṅka boldface variant of curve 

Seal-Id Inscription Sanskrit Translation

M-917a  श�ण śāṇa flaxen[ŚBr]

M-1903  अम�न� amān measure[√म� + लङ� 3p.]

 = अ · a from 1, 8, 27 (32)

1072 8.33.  · अ · a from अङ्क aṅka variant of curve 

Seal-Id Inscription Sanskrit Translation

M-1569  आश� āśā desire[AV]

M-1137  रव�ण ravāṇa roaring[√रु Ved]

 = अ · a from 1, 17, 19, 31 (33)

1073 8.34.  · अ · a from अङ्क aṅka variant of curve 

Seal-Id Inscription Sanskrit Translation

M-951  अम�न amāna not standard[Nyāyam]

M-1179  आन āna nose[RV]

 = अ · a from 1, 2, 16 (34)

62
1074 8.35.  · म · m from मन्थ mantha firesticks

Seal-Id Inscription Sanskrit Translation

M-501  अश्मक� aśmakam son of Vasiṣṭha[MBh]

M-257  दमन damana charioteer[BhP]

 = म · m from 1, 2, 3, 24, 25 (35)

1075 8.36.  · म · m from म�क्ष mṛkṣa variant of comb 

Seal-Id Inscription Sanskrit Translation

M-1989B  मन�क� manāk slightly[Kāv]

M-1233a  म�ण maṇi gem[RV]

 = म · m from 1, 5, 25, 26 (36)

1076 8.37.  · न · n from न�ल nāla mat of reeds

Seal-Id Inscription Sanskrit Translation

M-382A  आनन anana the living[Nir]

H-1682  जननजय janana-jaya creator[RV] of victory[AV]

 = न · n from 4, 11, 21 (37)

1077 8.38.  · व · v from वत� vartī lamp wick

Seal-Id Inscription Sanskrit Translation

M-925A  नवन navana praising[Nalod]

M-1689a  वश्य vaśya dutiful[MBh]
H-194  मवर� mavaraṃ multitudes[Buddh]

 = व · v from 1, 8, 13, 16, 21, 28 (38)

63
1078 8.39.  · श · s from श�खर śākhāra squirrel

Seal-Id Inscription Sanskrit Translation

H-1080a  श�ण śāṇa hempen[ŚBr]

M-1829a  श�न śani Saturn[R]
H-1830a  शय śaya sleeping[Dhātup]
L-35A  मसन masana transformation[√मस� + ल्य�ट]

 = श · s from 1, 5, 16, 21, 22 (39)

1079 8.40.  · म · m from म�न्दर mandira dwelling

Seal-Id Inscription Sanskrit Translation

M-832a  शमन śamana soothing[MBh]

H-892B  शमन� śamanī soothing[MBh]
H-854  जम��न jamāni may I eat[√जम� + ल�ट� 1s.]

 = म · m from 1, 5, 11, 39 (40)

1080 8.41.  · अ · a from अय�ग ayuga one

Seal-Id Inscription Sanskrit Translation

Banawali  आन āna nose[RV]

H-2570  आच� ācaṃ A name[Rājat]
M-1458  आनन�चर ā-anana-acara to[RV] constant[RV] living[Nir]

 = अ · a from 1, 2, 4, 13, 14, 23, 37 (41)

64
1081 8.42.  · र · r from रथद�रु rathadāru Dalbergia tree

Seal-Id Inscription Sanskrit Translation

H-1975A  कम� तर� kam taraṃ pleased[TS] Shiva[MBh]

H-289A  चरण caraṇa feet[MBh]
Dholavira  म�रण māraṇa destruction[Mn]

 = र · r from 1, 7, 8, 14, 25, 35, 40 (42)

1082 8.43.  · स · s from �शखर śikhara mountaintop

Seal-Id Inscription Sanskrit Translation

H-441a  सम� samaṃ level[RV]

H-201A  मसर� masaraṃ measurer[W,RV]

 = स · s from 1, 16, 40, 42 (43)

1083 8.44.  · ह · h from स=ह

Seal-Id Inscription Sanskrit Translation

M-1705  आमह� āmahaṃ to[RV] might[RV]

M-1845  आमह� āmahaṃ to[RV] might[RV]
M-445  आमह� āmahaṃ to[RV] might[RV]
H-474  असह�म� asahām impatient[Kathās]

 = ह · h from 1, 8, 24, 26, 27, 39, 43 (44)

65
1084 8.45.  · द · d from ध�नक�� dhānakāḥ variant of coins 

Seal-Id Inscription Sanskrit Translation

H-1740  दद dada given[√द� + �लट 2p.]

M-1088  रवद�न rava-dāna Roarer’s[√रु + अच� ] gift[RV]
M-1379  दद�न dadāna gave[√द�न� + �लट� 3s.]

 = द · d from 1, 6, 17, 19 (45)

1085 8.46.  · र · r from रथव� ratharvī multi/split snake

Seal-Id Inscription Sanskrit Translation

D-9093  हन�वचर hana-avacara killing[RV] domain[Buddh]

M’daro 84-2  वर vara gift[RV]
M-361  अरर arara door[Mcar]

 = र · r from 1, 4, 14, 17, 19, 39, 41 (46)

1086 8.47.  · य · y from यव yava barley

Seal-Id Inscription Sanskrit Translation

D-16261  आवय�जय āvaya-ajaya remove [√अज� + ल�ट� 2s.] pain[AV]

M-218A  यम� yamī (Yama’s) twin sister[RV]
H-48A  जय jaya victory[AV]
M-1322a  अजय ajaya undefeated[RV]

 = य · y from 2, 4, 5, 11, 15, 16, 17, 41 (47)

66
1087 8.48.  · द · d from ध�नक�� dhānakāḥ coins

Seal-Id Inscription Sanskrit Translation

M-619A  दह daha He who burns[√दह� + अच� RV]

H-844A  वदन vadana face[ŚBr]
M-190  सदस्य sadasya sacrifice attendee[TS]

 = द · d from 1, 17, 21, 31 (48)

1088 8.49.  · त · t from त�ड्य tāḍya drum

Seal-Id Inscription Sanskrit Translation

M-228A  मथन mathana churn[MBh]

M-459A  तन tana offspring[AV]
M-2079A  तद� tad there[AV]

 = त · t from 1, 16, 48 (49)

1089 8.50.  · म · m from मत� ग mataṅga elephant head

Seal-Id Inscription Sanskrit Translation

M-286a  आनमन ānamana saluting[T]

H-921A  त्मन� tman the self[RV]

 = म · m from 1, 7, 29 (50)

1090 8.51.  · त · t from त�ड� ल tāḍula fighter

Seal-Id Inscription Sanskrit Translation

M-1743A  श�न्तधर śānta-dhara tranquil[MBh] bearer[MBh]

H-1666A  तन�य tana-āya gain[MBh] of offspring[AV]

 = त · t from 1, 2, 4, 13, 21, 39, 48 (51)

67
1091 8.52.  · स · s from सप्तन� saptan seven

Seal-Id Inscription Sanskrit Translation

M-673  सर sara move[√स� + ल�ट� 2s.]

Krs-2  समन� saman keeping calm[√सम� + शत��]
H-9  सम sama equal[RV]
H-296  द�सम��न dāsamāni patronym द�सम�न�[√द�स� + श�नच� ]

 = स · s from 1, 5, 6, 13, 16, 36, 41 (52)

1092 8.53.  · र · r from रथ��रन� rathārin variant of wheel 

Seal-Id Inscription Sanskrit Translation

H-668A  रण raṇa joy[RV]

H-841A  ऋण� ṛṇī debtor[BG]
H-1951A  रज अश्नन� raja aśnan color[√रञ्ज� + ल�ट� 2s.]
while eating[√अश� + शत��]

 = र · r from 1, 5, 11, 24, 37 (53)

1093 8.54.  · श · s from श�खर śākhāra variant of squirrel 

Seal-Id Inscription Sanskrit Translation

H-771  दहस� dahas destroyer[√दह� + अच� RV]

M-1202A  दस्र�न्त�न� dasra-antān Dasra’s[RV] boundaries[RV]

 = श · s from 1, 22, 30, 31, 48, 53 (54)

68
1094 8.55.  · प · p from पञ्चन� pañcan hand

Seal-Id Inscription Sanskrit Translation

H-443A  अमपण ama-paṇa powerful[RV] wager[MBh]

H-758A  प�न pāna drink[RV]
M-967A  (अ)मपर� (a)ma-parī attain[AV] power[RV]
H-101a  पण� parṇa leaf[RV]
D-19655  यजपर yaja-para sacrifice[ŚBr] protector[√प� + अच� ]

 = प · p from 1, 2, 5, 9, 11, 12, 16, 22, 42, 47 (55)

1095 8.56.  · द · d from दन्त danta teeth

Seal-Id Inscription Sanskrit Translation

M-326c  दधद� dadhad wearing[√ध� + शत��]

H-2336  म�दन mādana exhilarating[RV]
H-86a  अदन adana food[RV]

 = द · d from 1, 2, 15, 16, 23, 48 (56)

1096 8.57.  · म · m from मय maya horse

Seal-Id Inscription Sanskrit Translation

H-2024  म� हन maṃhana gift[RV]

M-93  मन� जय manaṃ-jaya mind[RV] conqueror[RV]

 = म · m from 1, 4, 11, 16, 21, 39 (57)

69
1097 8.58.  · म · m from मत्य matya harrow

Seal-Id Inscription Sanskrit Translation

L-52a  आशम� āśam food[ŚBr]

M-965a  मनस्वर� manas-varaṃ willingly[RV] chosen[RV]

 = म · m from 1, 13, 24, 31, 38 (58)

1098 8.59.  · अ · a from आज�न ājani variant of stick 

Seal-Id Inscription Sanskrit Translation

M-950a  म�ख� mākhaṃ oblation[Hariv]

M-97A  म�य māya illusion[MBh]
M-189a  स�न sāna destruction[√स� + ल्य�ट�]
L-11a  रवस्रम आनव ravas-rama O pleasing[RV] Roarer[RV]
ānava O one kind to men[RV]

 = अ · a from 1, 17, 19, 21, 24, 25, 29, 38, 39, 42, 50 (59)

1099 8.60.  · ल · l from लत� latā creeper

Seal-Id Inscription Sanskrit Translation

M-13a  द�लन dālana decay[Suśr]

H-390a  अमलरम amala-rama pleasingly pure[AV,MBh]
M-183A  ल� मन laṃ mana greatly[Vā] pleasing[SBr 5.3.2.3]

 = ल · l from 3, 4, 9, 16, 19, 23, 27, 35, 58 (60)

70
1100 8.61.  · त · t from त�ल tāla small cymbal

Seal-Id Inscription Sanskrit Translation

K-120  जठर jaṭhara belly[RV]

 = त · t from 11, 13 (61)

1101 8.62.  · अन� · an from अ� श� aṃśu variant of drinking vessel 

Seal-Id Inscription Sanskrit Translation

H-450a  ज�नन� म�न jānan māna knowing[√�� + शत��] honored[MBh]

M-1787a  अन्वत� anvataṃ following[ĀpŚr]

 = अन� · an from 1, 7, 11, 15, 16, 17, 22, 26 (62)

1102 8.63.  · अ · a from अग aga mountain

Seal-Id Inscription Sanskrit Translation

H-139A  अरण araṇa foreign[RV]

H-157A  अन्य anya inexhaustible[AV]
M-1307a  अदन adana eating[√अद� + ल्य�ट�]

 = अ · a from 1, 9, 12, 21, 37, 56 (63)

71
1103 8.64.  · भ · b from भक्षत्र bhakṣatra oven

Seal-Id Inscription Sanskrit Translation

H-136A  मञ्जभर mañja-bhara shine[√मञ्ज� ] bestower[RV]

M-742A  आभ�म� सर�(म� ) ābhām sarā(m) river[AV] of light[MBh]
M-1777a  सभ�म� sabhām council[RV]
M-492A  महभण maha-bhaṇa mighty[RV] Roarer[√भण� + अच� ]

 = भ · b from 1, 2, 4, 8, 11, 12, 13, 16, 26, 29, 31, 35, 52 (64)

1104 8.65.  · ष · s from षण� ṣaṇ six

Seal-Id Inscription Sanskrit Translation

H-98A  समन्द्र sa-mandra with[RV] pleasantness[RV]

M-1314A  आनहर āna-hara (Sarasvati’s) nose[RV] severer[MBh]
H-514a  रवस्मय ravas-maya Roaring[RV,मयट� ]

 = ष · s from 1, 6, 12, 13, 16, 17, 19, 21, 23, 26, 41 (65)

1105 8.66.  · य · y from यम yama variant of yama 

Seal-Id Inscription Sanskrit Translation

M-1628  नयद nayada prudence[R] giver[MBh]

M-1628  रस्य rasya savoury[MBh]
M-1628  नस्य nasya nasal(breath)[ŚBr]

 = य · y from 6, 9, 19, 31, 37 (66)

72
1106 8.67.  · श · s from �शख� śikhā variant of peacock crest 

Seal-Id Inscription Sanskrit Translation

M-2033A  दशरथ्य daśa-rathya ten[RV] chariot horses[RV]

M-1096  अमम� श� amam-śam powerful[RV] blessing[RV]

 = श · s from 1, 6, 12, 16, 27, 51, 66 (67)

1107 8.68.  · उ · u from उ��म udyāma coil of rope

Seal-Id Inscription Sanskrit Translation

M-1773A  उदर� udaraṃ belly[RV]

Ad-5A  उदय udaya sunrise[ŚBr]

 = उ · u from 1, 21, 42, 56 (68)

1108 8.69.  · इ · i from इष�क iṣīkā variant of stalk of grass 

Seal-Id Inscription Sanskrit Translation

H-1511  अ�ह ahi Vrta[RV]

M-1632  ईम्प īmpa his[RV ईम� = एनम� ] protector[RV]
H-2102  ऐ ai vocative particle
Luristan  आ�द� ādiḥ the beginning[ChandUp]

 = इ · i from 4, 8, 16, 24, 48, 55 (69)

1109 8.70.  · व · v from वरट� varaṭī wasp

Seal-Id Inscription Sanskrit Translation

M-212A  व��णज vāṇija merchant[YV] (Shiva)

M-948A  अववज्य ava-vajya off[RV] journey[√वज� + ल्यप� ]

 = व · v from 1, 2, 11, 17, 47, 69 (70)

73
1110 8.71.  · क · k from क�तम� kṛtam axe

Seal-Id Inscription Sanskrit Translation

H-3a  अन्नकवष्सभ�� anna-kavaṣ- open[RV] food[RV]

sabhāṃ meeting[RV]
M-56A  धक्क� म�म� dhakkaḥ mām The destroyer[√धक्क� + अच� ]
सरन� saran moves[√स� + शत��] me
H-1076a  रव�ङ्कम� rava-aṅkam Roarer’s[√रु + अच� ] mark[R]
Dholavira  रकवर�रक- raka-varāraka- “gem of chosen gems”
Signboard अररस� araras entrance[Mcar]

 = क · k from 1,2,17, 18,19,24,25, 27,29,31,35, 37,48,52,53, 62,64,70 (71)

1111 8.72.  · क · k from क�ष kṛaṣa ploughshare

Seal-Id Inscription Sanskrit Translation

M-1889A  क�न्दर kāndara from a valley[R]

H-1987A  वङ्कश्चर vaṅkas-cara roam[Bhadrab] about[√चर� ]
M-1684a  अक्र akra banner[RV]

 = क · k from 1, 2, 13, 14, 38, 48, 67 (72)

1112 8.73.  · प · p from पञ्चन� pañcan five

Seal-Id Inscription Sanskrit Translation

M-1909A  अश्ववपन aśva-vapana cutting[ŚBr] the horse[RV]

M-1202C  आ-उप ā-upa from[RV] above[RV]

 = प · p from 2, 4, 9, 17, 26, 38, 48, 67 (73)

74
1113 8.74.  · उ · u from उप�नहन� upanihan hammer

Seal-Id Inscription Sanskrit Translation

M-1822  अन�ष� anu-uṣā post-dawn[Pāṇ,RV]

M-1180  य�प yūpa sacrificial post[RV]

 = उ��म · u from 1, 2, 21, 31, 73 (74)

1114 8.75.  · छ · c from छत्त्र chattra mushroom

Seal-Id Inscription Sanskrit Translation

H-1148  छन्न channa covered[MBh]

M-1129   दहनछद� dahana-chadaṃ cremation[BhP] shroud[BhP]
H-530a  अ�छन्न a-chinna undivided[�छद� + क्त]
H-642  जटध�चल� jaṭa-dhā-acalā immovable[BG] Shiva[Hariv,RV]

 = छ · c from 1, 2, 3, 4, 5, 6, 11, 23, 37, 41, 45, 60, 61, 67 (75)

1115 8.76.  · म · m from म�क्ष mṛkṣa comb

Seal-Id Inscription Sanskrit Translation

Ghola Dhoro  म�न्दद�चल mānda-acala gladdening[VS] mountain[MBh]

M-1955  रवस� ध� rava saṃdhṛ O Roarer[√रु], O bearer[MBh]

 = म · m from 1, 3, 4, 6, 8, 13, 17, 19, 26, 60, 75 (76)

75
1116 8.77.  · भ · b from भक्षपत्त्र� bhakṣapattrī betel leaf

Seal-Id Inscription Sanskrit Translation

H-1850A  भरण� bharaṇī N. stellar mansion[AV]

M-213A  मव� भ�� marva bhāṃ fill with[√मव� + ल�ट� 2s.] shining[RV]
H-5a  नट naṭa O dancer[MBh], O giver[Pāṇ] of
दहभस�नद daha-bhasāna-da heat[√दह� + अच� ] and shine[√भस� + ल्य�ट�]

 = भ · b from 1, 5, 6, 9, 13, 17, 24, 29, 31, 42, 51, 57 (77)

1117 8.78.  · अ · a from अङ्क aṅka variant of curve 

Seal-Id Inscription Sanskrit Translation

M-331  आभस� ā-bhasā the shining one[RV,√भस� ]

 = अ · a from 43, 77 (78)

1118 8.79.  · त · t from त�� tardū wooden ladles

Seal-Id Inscription Sanskrit Translation

M-1913  रवत ravata roaring[ĀpŚr]

M-741A  तय taya protector[√तय� + अच� ]

 = त · t from 12, 17, 21 (79)

1119 8.80.  · म · m from मन्द�र mandāra churning stick

Seal-Id Inscription Sanskrit Translation

M-1826a  दमन damana subduer[MBh]

M-1961a  रव मर rava mara O Roarer, O Killer[√रु + अच� ,√म� + अच� ]

 = म · m from 6, 12, 13, 17, 23 (80)

76
1120 8.81.  · क · k from क�ष kṛaṣa ploughshare

Seal-Id Inscription Sanskrit Translation

H-152a  क्षय kṣaya dominion[RV]

H-146  अक्षय akṣaya undecaying[BG]
M-1419  क्षय� kṣayī consumptive[MBh]

 = क · k from 2, 5, 8, 21 (81)

1121 8.82.  · ध · d from ध�न dhāna receptacle

Seal-Id Inscription Sanskrit Translation

BM-123208  छद�स्त chada-asta shroud[MBh,ŚBr]

Ch-5A  अनलवदत्त� ana-lava-dattaṃ breath[ŚBr]
destroyer’s[Mn] gift[RV]

 = ध · d from 1, 8, 17, 30, 41, 51, 60, 75 (82)

1122 8.83.  · झ · j from झञ्झ�न� jhañjān rain and wind

Seal-Id Inscription Sanskrit Translation

M-1308a  चर�जष� cara-a-jaṣaṃ moving[MBh] immortal[√जष� + अच� ]

H-57a  अज�य ajāya for Creator[RV]
M-836A  ज�न jani woman/wife[RV]
M-1848a  जष jaṣa killer[√जष� + अच� ]

 = झ · j from 1, 5, 9, 13, 14, 21, 26, 31, 41 (83)

77
1123 8.84.  · म · m from म�क्ष mṛkṣa variant of comb 

Seal-Id Inscription Sanskrit Translation

M-648A  आम अ� शम�न āma aṃśa-māna serve[√अम� + ल�ट� 2p.] portion[RV]

preparation[RV]
H-1706A  मल्लवज्र malla-vajra strong[MBh] Vajra[RV]
H-6A  भरस�दम� bhara-sādam carrying[RV] on horseback[RV]
K-40A  दमधरव�र dama-dhara-vāra householder’s[RV,RV]
treasure[RV]

 = म · m from 1,2,3, 4,11,12, 13,16,17, 20,22,26, 31,41,42, 48,60,82 (84)

1124 8.85.  · त · t from तद� tarda Indian blackbird

Seal-Id Inscription Sanskrit Translation

M-1204  मम अथ� mama athā mine certainly[Ved]

M-214  र�त ऐ rata ai O bestowed[RV]
M-1896  तत्र tatra there[RV]

 = त · t from 5, 12, 19, 26, 50, 78, 84 (85)

1125 8.86.  · म · m from मत्यय matya variant of churning stick 

Seal-Id Inscription Sanskrit Translation

M-779  दमन damana O Subduer

M-495  म�न्य�थ�दस mānya-atha-adas honored[RV] surely[RV,AV]

 = म · m from 1, 6, 21, 23, 26, 31, 48, 78, 85 (86)

78
1126 8.87.  · द · d from धन्वन� dhanvan variant of bow 

Seal-Id Inscription Sanskrit Translation

H-7a  दद एकपद� dada ekapadaḥ given[Pāṇ] by Viṣṇu[MBh]

M-159A  मदनह�न madana-hāna Kāma[MBh] remover[Gaut]
M-331a  अम�दभच�प ama-adabha-cāpa powerful[RV]
benevolent[RV] bow[MBh]

 = द · d from 1, 6, 14, 15, 20, 25, 26, 36, 41, 45, 52, 55, 73, 78, 80 (87)

1127 8.88.  · द · d from धन्वन� dhanvan variant of bow 

Seal-Id Inscription Sanskrit Translation

H-91  कम� दमन kam damana excellent[RV] subduer[MBh]

M’daro 84-6  दव dava heat[Car]

 = द · d from 23, 25, 35, 38, 80 (88)

1128 8.89.  · स · s from �शख� śikhā variant of 

Seal-Id Inscription Sanskrit Translation

M-1744  ध�मस��द� dhāma-sādi majestic[RV] horseman[MBh]

 = स · s from 2, 3, 5, 15, 27, 41, 87 (89)

1129 8.90.  · म · m from मन्द�र mandāra variant of flower 

Seal-Id Inscription Sanskrit Translation

M-810a  आमन āmana affection[TS]

M-308A  मम वमन� mama vaman my emitting[√वम� + शत��]

 = म · m from 1, 16, 17, 23, 26, 84 (90)

79
1130 8.91.  · म · m from म�न्दर mandira dwelling

Seal-Id Inscription Sanskrit Translation

D-24795  उन्नम� unnam moist[KātyŚr]

 = म · m from 1, 68 (91)

1131 8.92.  · त · t from त्र tra three

Seal-Id Inscription Sanskrit Translation

H-1711  क�न्तद�न kānta-dāna delightful[GargaS] gift[RV]

H-1713  रतजय rata-jaya pleasing[BhP] victory[RV]
H-351  तमस��न tamasāni dark-colored (pl)[AV]

 = त · t from 1, 2, 5, 11, 12, 31, 35, 47, 48, 72 (92)

1132 8.93.  · उ · u from उदप�न udapāna well

Seal-Id Inscription Sanskrit Translation

M-1923  श� अपरम� śu aparam the future[RV] quickly[Naigh]

M-1224  उस्र�� usreeḥ morning-lights[RV]
Unknown  उज्झमत� मन ujjha-matam mana remember,[√म्न� + ल�ट� 2s.]
relinquishing[Mn] desires[RV]

 = उ · u from 1, 4, 5, 11, 13, 16, 23, 42, 50, 55, 61, 65, 80 (93)

80
1133 8.94.  · उ · u from उदप�न udapāna variant of well 

Seal-Id Inscription Sanskrit Translation

H-720  उन्न unna moist[KātyŚr]

H-149  रवम� त�ञ्जतर ravam tuñja-tara Roarer, the attack protector[Nir]

 = उ · u from 1, 7, 11, 13, 16, 17, 19, 29, 49 (94)

1134 8.95.  · ए · e from एक eka one

Seal-Id Inscription Sanskrit Translation

M-1087A  अ�� षम� aśleṣam Hydra constellation[AV]

 = ए · e from 24, 58, 60 (95)

1135 8.96.  · अस� · as from अष्टन� aṣṭan eight

Seal-Id Inscription Sanskrit Translation

M-283A  अशर aśara unwounded[W]

H-818  अश्मन� aśman sky[RV]

 = अस� · as from 1, 13, 40 (96)

1136 8.97.  · न · n from नव nava nine

Seal-Id Inscription Sanskrit Translation

H-4a  अन्तम antama nearest[RV]

Dholavira  नय naya leading[R]
H-23a  नरह�न nara-hāna departed[VS] man[TS]

 = न · n from 1, 7, 13, 16, 21, 22, 26, 39 (97)

81
1137 8.98.  · र · r from रथ ratha variant of chariot 

Seal-Id Inscription Sanskrit Translation

H-24a  मञ्ज शरण mañja śaraṇa refuge[RV] O Roarer[√मञ्ज� + अच� ]

M-255a  मरण maraṇa death[MBh]
M-1366A  तर शतम� tara śatam a hundred[RV], O Rudra[MBh]

 = र · r from 1, 7, 8, 11, 16, 29, 39 (98)

1138 8.99.  · य · y from य�ष्ट yaṣṭi pearl necklace

Seal-Id Inscription Sanskrit Translation

H-465A  आय�न āyana arrival[RV]

M-324b  य�न yāna leading[RV]
M-99A  र�न्नव्य� rān-navyaṃ giving[√र� + शत��] praise[RV]

 = य · y from 1, 2, 4, 9, 12, 17 (99)

1139 8.100.  · म · m from म�क्ष mṛkṣa variant of comb 

Seal-Id Inscription Sanskrit Translation

H-1051  रव�समद्ससम��न ravā asamad-samānī O roarer, the calm[SBr]

and equable one[RV]

 = म · m from 1, 5, 6, 12, 17, 24, 36, 52 (100)

82
1140 8.101.  · य · y from य�ष्ट yaṣṭi twig; arm

Seal-Id Inscription Sanskrit Translation

M-290a  यम� yamī (Yama’s) twin sister[RV]

H-1922  असम�मम� य��न asama-amam yāni unequalled[Mn]
path[TS] of power[RV]

 = य · y from 1, 5, 16, 24, 27, 57, 100 (101)

1141 8.102.  · त · t from त�� tardū variant of wooden ladles 

Seal-Id Inscription Sanskrit Translation

H-90A  पश्यन्त� paśyantaḥ onlookers[RV]

C-6   आव आम�न तत्र āva āmāna tatra go[√अव� + ल�ट� 2p.] there[RV]
to[RV] honor[RV]

 = त · t from 1, 2, 4, 9, 12, 16, 17, 24, 26, 55, 63, 78, 101 (102)

1142 8.103.  · झ · j from झर jhara variant of rain; waterfall 

Seal-Id Inscription Sanskrit Translation

M-1691a  जन्मम�ख� janma-mākhaṃ birth[RV] oblation[Hariv]

K-30A  जज jaja warrior[Śiś]

 = झ · j from 1, 16, 25, 50, 59, 83 (103)

83
1143 8.104.  · ग · g from ग�धन gādhana arrow

Seal-Id Inscription Sanskrit Translation

H-1924A  गज� gajaḥ elephant[Mn]

M-1749a  गद�श्म gadā-aśma club[MBh] of stone[Pāṇ]
M-972a  ग�न gāna song[Śiś]
Allahdino-4A  घन�म�ह�� ghanā-māhāṃ great[MBh] destroyer[RV]

 = ग · g from 1, 2, 3, 4, 11, 16, 24, 27, 52, 63, 67 (104)

1144 8.105.  · म · m from मत्य matya club

Seal-Id Inscription Sanskrit Translation

M-1850a  रवनमन rava-namana bowing[Śiś] to Roarer[√रु + अच� ]

H-391  दम�मम�न dama-ama-māna this[RV] honored[MBh] house[RV]
M-285  रवमच्छम� ravam-accham towards[Ved] the Roarer[√रु + अच� ]

 = म · m from 1, 2, 6, 12, 16, 17, 27, 28, 35, 75 (105)

1145 8.106.  · म · m from मन्द�र mandāra a flower

Seal-Id Inscription Sanskrit Translation

H-1411  सम�न samāna uniform[RV]

H-950  शमशम śamaśama perpetually tranquil[MBh]
M-760  मनस्स्थम� manas-stham stilled[RV] mind[RV]

 = म · m from 1, 8, 16, 30, 31, 52, 67 (106)

84
1146 8.107.  · स · s from श�ण śāṇa weight of four

Seal-Id Inscription Sanskrit Translation

M-260A  रवसम�न rava-samāna like[RV] the Roarer[√रु + अच� ]

H-472a  रस�भ� rasa-ābhaṃ essence[RV] of beauty[MBh]

 = स · s from 1, 4, 12, 16, 17, 19, 26, 64 (107)

1147 8.108.  · अस� · as from अष्टप�द aṣṭapāda spider variant of 

Seal-Id Inscription Sanskrit Translation

M-628a  आशक लवमह�न� āśaka lava-mahān O bestower[RV]

O great reaper[MBh,RV]
H-1708A  असत्त्व asattva non-presence[Nyāyam]
M-1262a  बभ अश्नचर babha aśna-cara illumined[√भ� + �लट 2s.]
the moving[Mn] clouds[Naigh]

 = अस� · as from 1, 13, 14, 16, 17, 24, 25, 38, 41, 49, 60, 64 (108)

1148 8.109.  · त · t from तण्ड� ल taṇḍula rice plant

Seal-Id Inscription Sanskrit Translation

B-28  चलत calata move[√चल� + ल�ट� 2p.]

D-26514  यत�न�� yatānā[ḥ] efforts[RV]

 = त · t from 1, 4, 21, 60, 75 (109)

85
1149 8.110.  · स · s from �शखर śikhara mountaintop

Seal-Id Inscription Sanskrit Translation

Ns-79  वहस� vahas bearing[MBh]

M-2063  श�ण śāṇa hempen[ŚBr]

 = स · s from 1, 4, 17, 31 (110)

1150 8.111.  · स · s from �शखर śikhara variant of mountaintop 

Seal-Id Inscription Sanskrit Translation

M-264a  शहम� saham mighty[RV]

L-98A  सहत्मन� sahat-man mighty[RV]
M-1709a  रवमहम� rava-maham mighty[RV] Roarer[√रु + अच� ]

 = स · s from 1, 12, 16, 17, 23, 30, 91 (111)

1151 8.112.  · स · s from �शखर śikhara variant of mountaintop 

Seal-Id Inscription Sanskrit Translation

H-1920  अदशन� adaśan not ten[ŚBr]

H-2575  वश vaśa O power[AV]

 = स · s from 1, 3, 4, 70 (112)

1152 8.113.  · ल · l from लत� latā creeper

Seal-Id Inscription Sanskrit Translation

M-751A  ललक lalaka playful[W] happiness[ChUp]

Dholavira  कमल�स्र kamala-asra red[TS] blood[Ragh]

 = ल · l from 13, 24, 25, 35 (113)

86
1153 8.114.  · श · s from �शख� śikhā peacock crest

Seal-Id Inscription Sanskrit Translation

M-241a  सभ� दहन sabhaṃ dahana the council[RV],

O Rudra[MatsyaP]
M-1674a  शव� मनम�न śarva mana-māna O Shiva[AV],
O greatly[Vā] honored[MBh]

 = श · s from 1, 4, 9, 12, 16, 17, 31, 48, 64, 78 (114)

1154 8.115.  · म · m from मन्द�र mandāra variant of flower 

Seal-Id Inscription Sanskrit Translation

M-34A  तम्रवस्मन� tamra-vasman red[VS] covering[RV]

M-1823A  अम�वचर ama-avacara domain[Buddh] of power[RV]
M-43a  मज्ज व मह�न� majja va mahān submerge[MBh] like
the great one[MBh]

 = म · m from 1, 2, 13, 14, 16, 17, 24, 34, 39, 40, 83, 109 (115)

1155 8.116.  · व · v from वट� vaṭī banyan tree

Seal-Id Inscription Sanskrit Translation

M-1159  स्ववश� नमन svavaśaṃ namana[ṃ] willing[MBh] respects[MārkP]

M-101  रव वन�र rava vānara O Roarer[√रु + अच� ],
O forest dweller[Mn]

 = व · v from 1, 12, 13, 16, 17, 28, 31, 63 (116)

87
1156 8.117.  · र · r from रथद�रु rathadāru variant of Dalbergia 

Seal-Id Inscription Sanskrit Translation

M-1773  उदर� udaraṃ belly[RV]

M-1514  रण raṇa joy[RV]
M-1275  अहसम� परम� ahasam param first[MBh] day[RV]

 = र · r from 1, 16, 24, 31, 55, 56, 68 (117)

1157 8.118.  · स · s from श्य� न śyena falcon

Seal-Id Inscription Sanskrit Translation

M-222a  मयस� mayas delight[RV]

M-6  सन्नरर san-arara existing[√अस� + शत��] door[Mcar]

 = स · s from 9, 12, 19, 50, 101 (118)

8.119. · त · t from त�ल tāla variant of cymbal 



1158

Seal-Id Inscription Sanskrit Translation

M-2118  हय च पत haya ca pata roarer[√हय� + अच� ] and master[√पत� + अच� ]



= त · t from 8, 14, 21, 73 (119)



1159 8.120.  · अस� · as from अष्टप�द aṣṭapāda variant of spider 

Seal-Id Inscription Sanskrit Translation

M-976  रण�क्ष raṇa-akṣa eye[RV] of Roarer[Pur]

 = अस� · as from 9, 12, 25, 41 (120)

88
1160 8.121.  · अस� · as from अष्टप�द aṣṭapāda variant of spider 

Seal-Id Inscription Sanskrit Translation

M-1975  द�सव�रम� dāsa-varam dāsa[RV] restrainer[RV]

 = अस� · as from 1, 6, 38, 41, 117 (121)

1161 8.122.  · ओ · o from ओपश opaśa tuft of hair

Seal-Id Inscription Sanskrit Translation

H-1924  ओजस� ojas vigour[RV]

 = ओ · o from 11, 24 (122)

1162 8.123.  · र · r from रथद�रु rathadāru variant of Dalbergia 

Seal-Id Inscription Sanskrit Translation

H-1522  ररर ra-ra-ra possessing[Naiṣ]

love and desire
L-45  तन�मन�मन� tana-ama-nāman fulfill[प� + ल�ट� 2s.] the mark[RV]
पर para of the destroyer[√तन� + अच� ]

 = र · r from 1, 16, 27, 28, 49, 55 (123)

1163 8.124.  · व · v from व�त्र vātra loom √ व� + ष्ट्रन�

Seal-Id Inscription Sanskrit Translation

M-899  व्यय vyaya sacrifice[R]

 = व · v from 21 (124)

89
1164 9. Derivation Sequence

Phoneme Reconstructed Name

1  अन� an अ� श� aṃśu soma drink

2  अ a आज�न ājani stick
3  द d धन्वन� dhanvan bow
4  अ a आय� āyu man
5  इ i इष�क iṣīkā stalk of grass
6  द d धन्वन� dhanvan variant of bow 
7  त t त�ड� ल taḍula variant of fighter 
8  श s श�क्र śukra seed
9  न n न�ल nāla stalk
10  ई ī इइ ī long i
11  ज j झर jhara waterfall; cascade
12  र r रथ ratha chariot
13  र r रथव� ratharvī multi/split snake
14  च c चत�र� catur four
15  अ a अजशृङ्ग� ajaśṛṅgī goat’s horn
16  म m मत्स्य matsya fish
17  व v �व vi two
18  र r रथ ratha variant of ratha 
19  र r रथ��रन� rathārin chariot wheel
20  ब b भक्षपत्त्र� bhakṣapattrī betel leaf
21  य y यव�� �� yavaśreṣṭhi grain merchant
22  अ a आय� āyu variant of 
23  न n न�ल�क� nālīka arrow
24  अस� as अष्टप�द aṣṭapāda eight legged; spider
25  क k क�तम� kṛtam dice
26  अ a अङ्क aṅka curve
27  अम ama अङ्क + मत्स्य ama conjunct curve + fish
28  न n न�ल nāla reed
29  अन� n अ� श� aṃśu lamp
30  त t त�� tardū variant of wooden ladles 
31  स s स�प�न sopāna ladder
32  अ a अङ्क aṅka boldface variant of curve 
33  अ a अङ्क aṅka variant of curve 
34  अ a अङ्क aṅka variant of curve 
35  म m मन्थ mantha firesticks
36  म m म�क्ष mṛkṣa variant of comb 
37  न n न�ल nāla mat of reeds
38  व v वत� vartī lamp wick
39  श s श�खर śākhāra squirrel
40  म m म�न्दर mandira dwelling
41  अ a अय�ग ayuga one
42  र r रथद�रु rathadāru Dalbergia tree
43  स s �शखर śikhara mountaintop

90
 Phoneme Reconstructed Name

44  ह h स=ह
45  द d ध�नक�� dhānakāḥ variant of coins 
46  र r रथव� ratharvī multi/split snake
47  य y यव yava barley
48  द d ध�नक�� dhānakāḥ coins
49  त t त�ड्य tāḍya drum
50  म m मत� ग mataṅga elephant head
51  त t त�ड� ल tāḍula fighter
52  स s सप्तन� saptan seven
53  र r रथ��रन� rathārin variant of wheel 
54  श s श�खर śākhāra variant of squirrel 
55  प p पञ्चन� pañcan hand
56  द d दन्त danta teeth
57  म m मय maya horse
58  म m मत्य matya harrow
59  अ a आज�न ājani variant of stick 
60  ल l लत� latā creeper
61  त t त�ल tāla small cymbal
62  अन� an अ� श� aṃśu variant of drinking vessel 
63  अ a अग aga mountain
64  भ b भक्षत्र bhakṣatra oven
65  ष s षण� ṣaṇ six
66  य y यम yama variant of yama 
67  श s �शख� śikhā variant of peacock crest 
68  उ u उ��म udyāma coil of rope
69  इ i इष�क iṣīkā variant of stalk of grass 
70  व v वरट� varaṭī wasp
71  क k क�तम� kṛtam axe
72  क k क�ष kṛaṣa ploughshare
73  प p पञ्चन� pañcan five
74  उ u उप�नहन� upanihan hammer
75  छ c छत्त्र chattra mushroom
76  म m म�क्ष mṛkṣa comb
77  भ b भक्षपत्त्र� bhakṣapattrī betel leaf
78  अ a अङ्क aṅka variant of curve 
79  त t त�� tardū wooden ladles
80  म m मन्द�र mandāra churning stick
81  क k क�ष kṛaṣa ploughshare
82  ध d ध�न dhāna receptacle
83  झ j झञ्झ�न� jhañjān rain and wind
84  म m म�क्ष mṛkṣa variant of comb 
85  त t तद� tarda Indian blackbird
86  म m मत्यय matya variant of churning stick 
87  द d धन्वन� dhanvan variant of bow 
88  द d धन्वन� dhanvan variant of bow 
89  स s �शख� śikhā variant of 

91
 Phoneme Reconstructed Name

90  म m मन्द�र mandāra variant of flower 

91  म m म�न्दर mandira dwelling
92  त t त्र tra three
93  उ u उदप�न udapāna well
94  उ u उदप�न udapāna variant of well 
95  ए e एक eka one
96  अस� as अष्टन� aṣṭan eight
97  न n नव nava nine
98  र r रथ ratha variant of chariot 
99  य y य�ष्ट yaṣṭi pearl necklace
100  म m म�क्ष mṛkṣa variant of comb 
101  य y य�ष्ट yaṣṭi twig; arm
102  त t त�� tardū variant of wooden ladles 
103  झ j झर jhara variant of rain; waterfall 
104  ग g ग�धन gādhana arrow
105  म m मत्य matya club
106  म m मन्द�र mandāra a flower
107  स s श�ण śāṇa weight of four
108  अस� as अष्टप�द aṣṭapāda spider variant of 
109  त t तण्ड� ल taṇḍula rice plant
110  स s �शखर śikhara mountaintop
111  स s �शखर śikhara variant of mountaintop 
112  स s �शखर śikhara variant of mountaintop 
113  ल l लत� latā creeper
114  श s �शख� śikhā peacock crest
115  म m मन्द�र mandāra variant of flower 
116  व v वट� vaṭī banyan tree
117  र r रथद�रु rathadāru variant of Dalbergia 
118  स s श्य� न śyena falcon
त t त�ल tāla variant of cymbal 


119

120  अस� as अष्टप�द aṣṭapāda variant of spider 

121  अस� as अष्टप�द aṣṭapāda variant of spider 
122  ओ o ओपश opaśa tuft of hair
123  र r रथद�रु rathadāru variant of Dalbergia 
124  व v व�त्र vātra loom √व� + ष्ट्रन�

92
1165 10. Conclusion

1166 The ability to read well beyond the unicity distance alone should be sufficient proof
1167 of correct decipherment of the Indus script. This decipherment has many additional
1168 compelling attributes. This is the only cryptanalytic decipherment and the only one
1169 that uses well-established mathematical models and methods instead of guessing sign
1170 values based on their appearance. This decipherment is the only full decipherment
1171 and the only one where every sign and every stroke has been resolved, the only one
1172 that is programmatically reproducible, the only one where the decipherment can be
1173 followed sign-by-sign by the reader, the only decipherment that reads Semitic and mixed
1174 inscriptions in addition to native IVC inscriptions, the only one that reads over 500
1175 inscriptions including all 50 longest inscriptions grammatically correctly in an attested
1176 language, the only one that validates research spanning almost a century from Hunter to
1177 Heggarty. In addition, we have uncovered a remarkable number of additional evidence
1178 such as reconstructed names of the signs, reasons for their allographs, and the clear
1179 correspondence of derived sound values to known Brahmi values. We also show how
1180 the constraints and habits of the Indus script carry on to Brahmi inscriptions of the
1181 early historic era. Such a strong result is a first in any ancient script decipherment and
1182 should be taken as plenary proof of decipherment of the Indus script.

1183 11. Data Availability Statement

1184 A programmatic decipherment of the first 40 signs is openly available in the GitHub
1185 repository at https://github.com/yajnadevam/ScriptDerivation. This paper uses the
1186 Indus script font from the National Fund for Mohenjodaro under open license(Kumb-
1187 har and Buriro, 2017). Brahmi and Devanagari fonts are from Google Fonts under Open
1188 Font license(Google, 2021, 2022). Adinata font for Tamil Brahmi is under Open Font
1189 license (Rajan, Sharma, and Sankar, 2021). The Indus corpus reference used is Interac-
1190 tive Corpus of Indus Text(Wells and Fuls, 2023). Corpus of Indus seals and inscriptions
1191 volumes are also the primary reference(Parpola et. al., 1991). Rigveda translations in
1192 section 5 are from Griffith(Griffith, 1896). The dictionary used for decipherment deriva-
1193 tion is the downloadable Monier-Williams dictionary(Monier-Williams, 1899b). Attes-
1194 tation data of individual words are from Monier-Williams dictionary(Monier-Williams,
1195 1899b), the Purana Index(Dikshitar, 1955) and Wisdomlib(Hiemstra, 2023).

1196 12. Disclosure Statement

1197 The authors report there are no competing interests to declare.

References

Ahmed, Mukhtar. 2014. Ancient Pakistan - An Archaeological History: Volume II: A

Prelude to Civilization. Foursome Group. https://books.google.com/books?id=
HbvTBAAAQBAJ&pg=PA364#v=onepage&q&f=false.
Aniket Anand, Deshmukh, and Soumya Jana. 2013. “Chronology of Sanskrit texts:
An information-theoretic corroboration.” In 2013 National Conference on Commu-
nications (NCC), 1–5. https://web.archive.org/web/20160806054822id_/http:
//www-personal.umich.edu:80/~aniketde/NCC%20paper.pdf.

93
Ashraf, M, and S Sinha. 2018. “The “handedness” of language: Directional symme-
try breaking of sign usage in words.” PLOS one 1–16. https://doi.org/10.1371/
journal.pone.0190735.
ASI. 1906. Archaeological Survey Of India Annual Report 1903-4. Government Print-
ing, India. https://archive.org/details/in.ernet.dli.2015.207529/page/
n195/mode/2up?view=theater.
ASI. 2004. Indian Archaeology 1998-1999 - A review. Archaeological Survey of In-
dia. https://web.archive.org/web/20120508064754/http://asi.nic.in/nmma_
reviews/Indian%20Archaeology%201998-99%20A%20Review.pdf.
Association Assyrophile de France. 2006. “Sureth Dictionary.” https://www.
assyrianlanguages.org/akkadian/list.php.
Azad, Uma Shankar. 2020. “Seals and Sealings of Pataliputra (From the Maurya Period
up to the Gupta Period).” Heritage: Journal of Multidisciplinary Studies in Ar-
chaeology 632–647. http://www.heritageuniversityofkerala.com/JournalPDF/
Volume8.2/37.pdf.
Bacon, Roger. 1401-1599. “Voynich manuscript.” https://collections.library.
yale.edu/catalog/2002046?child_oid=1006204.
Balasubramaniam, R. 2005. Story of the Delhi Iron pillar. Foundation Books pvt Ltd.
https://books.google.com/books?id=FruHOlKlbJAC.
Banerjee, N. R., and Soundara K. V. Rajan. 1960. “Sanur 1950, 1952: A megalithic
site in district Chingleput.” Ancient India 15: 2–28. https://asi.nic.in/Ancient_
India/Ancient_India_Volume_15/article_1.pdf.
BBC. 1999. “’Earliest writing’ found.” Last accessed 18 March 2022, https:
//web.archive.org/web/20220318234228/http://news.bbc.co.uk/2/hi/sci/
tech/334517.stm.
Bhatt, S. K. 1998. Nishka-the Rig Vedic Money. Academy of Indian Numismatics and
Sigillography.
Bonta, Steven. 2023. “A Partial Decipherment of the Indus Valley Script: Proposed
Phonetic and Logographic Values for Selected Indus Signs and Readings of Indus
Texts.” academia.edu https://web.archive.org/web/20230913171235/https:
//www.academia.edu/105163460/A_Partial_Decipherment_of_the_Indus_
Valley_Script_Proposed_Phonetic_and_Logographic_Values_for_Selected_
Indus_Signs_and_Readings_of_Indus_Texts.
CDLI. 2023. “The CDLI Collection.” https://cdli.ucla.edu/.
Cinque, Guglielmo. 1999. Adverbs and Functional Heads. Oxford University Press.
https://www.academia.edu/58433772/Adverbs_and_Functional_Heads.
CoinIndia. 2010. “The Coin Galleries: Mauryan Empire.” http://coinindia.com/
galleries-maurya.html.
Danino, Michael. 2006. “THE HORSE AND THE ARYAN DEBATE.” Journal of Indian
History and Culture September 2006 (13): 33–59. https://www.researchgate.net/
publication/237413669_THE_HORSE_AND_THE_ARYAN_DEBATE.
Devi, Yashoda. 1933. The history of Andhra country (1000 AD - 1500 AD). Gyan
Publishing House.
Dikshitar, Ramachandra. 1955. “The Purana Index.” https://www.
sanskrit-lexicon.uni-koeln.de/scans/PUIScan/2020/web/index.php.
Dixit, K. N. 2019. “Origin of Early Harappan Cultures in the Sarasvati
Valley: Recent Archaeological Evidence and Radiometric Dates.” Jour-
nal of Indian Ocean Archaeology 88–141. https://web.archive.org/
web/20170118032736/http://server2.docfoc.com/uploads/Z2015/11/21/
vESLakMBYz/45a03572f94e7a873d7c350293cca188.pdf.

94
Ephron, H. D. 1961. “MYCENAEAN GREEK: A LESSON IN CRYPTANALYSIS.”
Minos: Revista de Filología Egea 63–100. https://revistas.usal.es/dos/index.
php/0544-3733/article/view/2088/2140.
Gelb, I. J. 1957. Glossary of old Akkadian. University of Chicago Press. https://isac.
uchicago.edu/sites/default/files/uploads/shared/docs/mad3.pdf.
George, Andrew. 2003. “CUNEIFORM INSCRIPTION OF DEFEAT OF
OMAN and INDUS VALLEY.” https://www.schoyencollection.com/
history-collection-introduction/sumerian-history-collection/
cuneiform-indus-valley-ms-2814.
George Lasry, Norbert Biermann, and Satoshi Tomokiyo. 2023. “Deciphering Mary
Stuart’s lost letters from 1578-1584.” Cryptologia 47 (2): 101–202. https://-
doi.org/10.1080/01611194.2022.2160677, https://doi.org/10.1080/01611194.
2022.2160677.
Goibhniu. 2007. “Perl uses for Cryptograms - Part 1: One-liners and
Word Patterns.” https://web.archive.org/web/20170223171834/https://www.
perlmonks.org/?node_id=636818.
Google. 2021. “Noto Serif Devanagari font.” This work is licensed under the
Open Font Licence. To view a copy of this license, visit https://scripts.sil.
org/cms/scripts/page.php?site_id=nrsi&id=OFL, https://fonts.google.com/
noto/specimen/Noto+Serif+Devanagari/about.
Google. 2022. “Noto Sans Brahmi font.” This work is licensed under the Open
Font Licence. To view a copy of this license, visit https://scripts.sil.
org/cms/scripts/page.php?site_id=nrsi&id=OFL, https://fonts.google.com/
noto/specimen/Noto+Sans+Brahmi/about.
Griffith, Ralph T. H. 1896. The Hymns of the Rigveda. Munshiram Manoharlal Publ.
Vedaweb version https://vedaweb.uni-koeln.de/rigveda/view/index/0, http://www.
sanskritweb.net/rigveda/griffith.pdf.
Haryana Directorate of Archaeology and Museums. 2023. “Excavation Sites A and M.”
https://web.archive.org/web/20230330043817/http://archaeologyharyana.
nic.in/sites/default/files/Excavation%20sites%20A%20and%20M.pdf.
Heggarty, Paul, Cormac Anderson, Matthew Scarborough, Benedict King, Remco
Bouckaert, Lechosław Jocz, Martin Joachim Kümmel, et al. 2023. “Language trees
with sampled ancestors support a hybrid model for the origin of Indo-European
languages.” Science 381 (6656): eabg0818. https://doi.org/10.1126/science.abg0818,
https://www.science.org/doi/abs/10.1126/science.abg0818.
Hiemstra, Gabe. 2023. “Wisdom Library.” https://wisdomlib.org.
Hultzsch, E. 1925. Inscriptions of Asoka. Clarendon Press. https://archive.org/
details/InscriptionsOfAsoka.NewEditionByE.Hultzsch/page/n110/mode/1up?
view=theater.
Hunter, G. R. 1934. The script of Harappana and Mohenjodaro and its connection
with other scripts. Kegan Paul, Trench, Trubner & Co Ltd. https://ufdc.ufl.edu/
AA00013642/00001.
IndMuseum. 5th Century. “Terracotta sealing.” https://indianculture.gov.in/
artefacts-museums/sealing-397.
Jayaswal, K. P. 1933. “The Vikramkhol inscription.” Indian Antiquary: A Journal Of
Oriental Research 62: 58–60. https://archive.org/details/in.ernet.dli.2015.
537235.
Kak, Subhash. 1988. “A Frequency analysis of the Indus script.” Cryptologia 129–142.
https://www.ece.lsu.edu/kak/IndusFreqAnalysis.pdf.
Knight, Kevin, Beáta Megyesi, and Christiane Schaefer. 2011. “The Copiale Cipher.”

95
 Association for Computational Linguistics 2–9. https://web.archive.org/web/
20111112013300/http://www.aclweb.org/anthology-new/W/W11/W11-12.pdf#
page=12.
Konasukawa, Ayumu. 2020. “https://www.youtube.com/watch?v=UWNzMDfN28I.” .
Kumbhar, Shabir, and Amar Fayaz Buriro. 2017. “Indus script font.” License ”Indus
Script font is available to be downloaded for further studies, computational exercises
and statistical analysis, free of charge; the only encumbrance is that user acknowl-
edge our website.”, https://web.archive.org/web/20210308035522/http://www.
mohenjodaroonline.net/index.php/indus-script/corpus-by-asko-parpola.
Lahafian, Jamal. 2013. “Rock Art in Kurdistan Iran.” https://www.researchgate.
net/publication/276033365_Rock_Art_in_Kurdistan_Iran.
Lal, B. B. 1960. “From the Megalithic to the Harappa: tracing back the grafitti on the
pottery.” Ancient India 16: 4–24. https://asi.nic.in/Ancient_India/Ancient_
India_Volume_16/article_1.pdf.
Lashari, K., Sindh (Pakistan). Department of Culture, Tourism, and Sindh Archives.
2020. Studies on Indus Cript: Conference on Indus Script Mohenjodaro 2020. National
Fund for Mohenjodaro, Culture, Tournism, Antiquities and Archives Department,
Government of Sindh. https://books.google.com/books?id=gjWCzwEACAAJ.
Laursen, Steffen Terp. 2010. “The westward transmission of Indus Valley seal-
ing technology: origin and development of the ‘Gulf Type’ seal and other ad-
ministrative technologies in Early Dilmun, c.2100–2000 BC.” Arabian archaeol-
ogy and epigraphy 96–134. https://onlinelibrary.wiley.com/doi/10.1111/j.
1600-0471.2010.00329.x.
Lien, L. T. 2013. “Brahmi Inscription from Archaeological Sites in South-
ern Vietnam during the 1st Millennium CE.” Advancing Southeast Asian Ar-
chaeology 402. https://publications.spafajournal.org/index.php/spafapub/
catalog/view/37/53/158-1.
Mahadevan, Iravatham. 1977. The Indus Script. Archaeological Sur-
vey of India. https://archive.org/34/items/TheIndusScript.
TextConcordanceAndTablesIravathanMahadevan/The%20Indus%20Script.
%20Text%2C%20Concordance%20and%20Tables%20-Iravathan%20Mahadevan.pdf.
Mahadevan, Iravatham. 2010. “Akam and Puram : ‘Address’ Signs of the Indus Script.”
https://rmrl.in/wp-content/uploads/42-Akam-and-Puram.pdf.
MetMuseum. 6th CE. “Impression from a Property Seal.” https://www.metmuseum.
org/art/collection/search/77782.
Monier-Williams. 1899a. “Monier-Williams dictionary Sanskrit abbreviations.” https:
//en.wiktionary.org/wiki/Appendix:Sanskrit_abbreviations.
Monier-Williams. 1899b. “Monier-Williams Sanskrit-English Dictionary, 1899.” https:
//www.sanskrit-lexicon.uni-koeln.de/scans/MWScan/2020/web/index.php.
Moreno, Marsha Lynn. 2005. “Frequency Analysis in Light of Language In-
novation.” https://mathweb.ucsd.edu/~crypto/Projects/MarshaMoreno/
TimeComparisonFrequency.pdf.
Munjal, Sanjay, and Arvin Munjal. 2005. “(Brahmi inscription on Anthropomorphic
figure).” Prāgdhārā .
Neumayer, Erwin. 2020. “Chariots in the Chalcolithic Rock Art of Indian.”
https://www.harappa.com/sites/default/files/pdf/WHEELS%20in%20Indian%
20Rock%20Art%20Erwin%20Neumayer.pdf.
Parpola, Asko. 1994. Deciphering the Indus script. Cambridge University Press.
Parpola, Simo, Asko Parpola, and Robert H. Brunswig Jr. 1977. “The Meluḫḫa Vil-
lage: Evidence of Acculturation of Harappan Traders in Late Third Millennium

96
 Mesopotamia?” Journal of the Economic and Social History of the Orient 20 (2):
129–165. http://www.jstor.org/stable/3631775.
Parpola et. al. 1991. Corpus of Indus seals and inscriptions, Vol 1,2,3.1,3.2,3.3. Suo-
malainen Tiedeakatemia.
Peters et al. 2022. “The biocultural origins and dispersal of domestic chickens.” PNAS
119 (24). https://www.pnas.org/doi/pdf/10.1073/pnas.2121978119.
Raghupathy, P. 1987. Early settlements in Jaffna: An archaeological survey. P. Raghu-
pathy.
Rajan, Vinod, Shriramana Sharma, and Udhaya Sankar. 2021. “Adinatha Tamil-
Brahmi font.” This work is licensed under the Open Font Licence. To view a copy
of this license, visit https://scripts.sil.org/cms/scripts/page.php?site_id=
nrsi&id=OFL, http://www.virtualvinodh.com/projects/adinatha.
Ramakrishna, K. A., N. K Swain, M. Rajesh, and N. Veeraraghavan. 2018. “Ex-
cavations at Keeladi, Sivaganga District, Tamil Nadu (2014 - 2015 and 2015 -
16).” Heritage: Journal of Multidisciplinary Studies in Archaeology 6 30–72. http:
//www.heritageuniversityofkerala.com/JournalPDF/Volume6/2.pdf.
Rao, Rajesh P. N., Nisha Yadav, Mayank N. Vahia, Hrishikesh Joglekar,
R. Adhikari, and Iravatham Mahadevan. 2009. “A Markov model of the
Indus script.” Proceedings of the National Academy of Sciences 106 (33):
13685–13690. https://doi.org/10.1073/pnas.0906237106, https://www.pnas.org/
doi/abs/10.1073/pnas.0906237106.
Rao, S. R. 1980. “Indus script and language.” Annals of the Bhandarkar Oriental Re-
search Institute 61: 157–188.
Reddy, Bhaskar, and V. Sakunthalamma. 2023. “Indian Numismatics:
Punchmarked Coins of the Janapadas.” https://epgp.inflibnet.ac.
in/epgpdata/uploads/epgp_content/S000829IC/P001738/M028921/ET/
1523266853P09-M04-PunchMarkedCoinsoftheJanapadas-ET.pdf.
Shannon, C. 1945. A Mathematical Theory of Cryptography. Alcatel-Lucent. https:
//www.iacr.org/museum/shannon/shannon45.pdf.
Sheehan, John. 2009. “‘A peacock’s tale: excavations at Caherlehillan, Iveragh,
Ireland.” The Archaeology of the Early Medieval Celtic Churches 191–206.
https://www.academia.edu/3123538/The_Peacock_s_Tale_excavations_at_
Caherlehillan_Kerry_Ireland.
Shishlina, N. I., L. N. Koryakova, and O. V. Orfinskaya. 2022. “Exotic Cotton Tex-
tile of the Bronze Age from the Southern Trans-Urals.” Nanobiotechnology Reports
17 (5): 691–700. https://www.academia.edu/94620659/Exotic_Cotton_Textile_
of_the_Bronze_Age_from_the_Southern_Trans_Urals.
Singh, Unpinder. 2006. Delhi: Ancient History. Berghahn Books. https://en.
wikipedia.org/wiki/Iron_pillar_of_Delhi.
Sinha, B.P., and Sita Ram Roy. 1969. Vaisali excavations (1958 - 1962). Directorate of
Archaeology and Museums. http://asi.nic.in/asi_books/71048.pdf.
Sivanantham, Dr. R., and M. Seran. 2019. Keeladi: An Urban Settlement of Sangam
Age on the banks of river Vaigai. Department of Archaeology, Government of Tamil
Nadu. https://web.archive.org/web/20231015164905/https://www.vinavu.
com/wp-content/uploads/2019/09/Keeladi-Book-English-18-09-2019.pdf.
Solomon, Richard. 1998. Indian Epigraphy. Oxford University Press.
Srivatsava, O. 2021. (New Dimensions of History and Archeology). B. R. Publishing
Corporation.
Subrahmanian, T. S. 2010. “Tamil Brahmi potsherds found at urn burial site.”
https://www.thehindu.com/features/friday-review/history-and-culture/

97
 Tamil-Brahmi-potsherds-found-at-urn-burial-site/article16483663.ece.
Sullivan, Sue. 2011. Indus script dictionary. Sullivan, Sue.
UPenn. 2006. “The Pennsylvania Sumerian Dictionary.” http://psd.museum.upenn.
edu/nepsd-frame.html.
Vyas, Shail. 2020. “Indus Musicians in Mesopotamia.” https://osf.io/preprints/
socarxiv/kce5x/download&usg=AOvVaw228Fmhab2qMXq19nY5AenT&opi=89978449.
Wells, Bryan, and Andreas Fuls. 2023. “Interactive Concordance of Indus Texts (ICIT):
An Online Database of Indus Inscriptions and Iconography.” https://www.indus.
epigraphica.de.
wikipedia. 2021. “List of Beijing Subway stations.” https://en.wikipedia.org/wiki/
List_of_Beijing_Subway_stations.

98
Appendix A. Evolution of Brahmi signs

Figure A1. Evolution of Brahmi signs: Wikimedia Commons

99