University of Nebraska - Lincoln

DigitalCommons@University of Nebraska - Lincoln

Faculty Publications, Classics and Religious Studies
Classics and Religious Studies
Department

January 1999

Roberto Busa, S.J., and the Invention of the

Machine-Generated Concordance
Thomas Nelson Winter
University of Nebraska-Lincoln, [email protected]

Follow this and additional works at: http://digitalcommons.unl.edu/classicsfacpub

Part of the Classics Commons

Winter, Thomas Nelson, "Roberto Busa, S.J., and the Invention of the Machine-Generated Concordance" (1999). Faculty Publications,
Classics and Religious Studies Department. 70.
http://digitalcommons.unl.edu/classicsfacpub/70

This Article is brought to you for free and open access by the Classics and Religious Studies at DigitalCommons@University of Nebraska - Lincoln. It
has been accepted for inclusion in Faculty Publications, Classics and Religious Studies Department by an authorized administrator of
DigitalCommons@University of Nebraska - Lincoln.
 ROBERTOBUSA, S.J., AND THE INVENTION
OF THE MACHINE-GENERATED CONCORDANCE1

"The use of the latest data-processing tools developed primarily for

science and commerce may prove a significant factor in facilitating future
literary and scholarly studies."
- Paul Tasman
IBM Journal of Research and Development (July, 1957), p. 249.

H
ere in 1999, Classicists, both as teachers and as scholars, are
equipped to appreciate that Tasman's 1957 prophecy was, if
anything, an understatement.
Classicists, both as teachers and as scholars, are enjoying the fruits
. . -

of the ongoing technological revolution. Classics teachers can now, with

a few keystrokes, produce a customized concordanceto any literary work
we teach; we can generate word-lists for our students in frequency-of-
occurrence order for efficiency-guided vocabulary study. The Internet
has become a world-wide burgeoning encyclopedia at our keyboard: we
can pull in texts off the Internet--or scan them ourselves-and run search
packages on them; we can pull in photos of vases, or of archaeological sites
and with a few more keystrokes turn them directly into transparencies for
classroom use.
Classicists, as well as any, know that Tasman got it right.
Given the advances in these "data-processing tools developed
primarily for science and commerce," as well as the work of Ted runner
and Lucy Berkowitz (the TLG, the TLG Canon) and of David Packard (the
concordance,Ibycus, the ~ackardHumanities Institute) and
Stephen Van Fleet Waite (LOGO1 Systems, now in PHI), research work
which even in 1980 required a mainframe computer can now be done at a
desktop. In sum, the history of computers and the humanities has been all
along a matter of broad-ranging scholars seeing opportunities in the
newest commercially available data-handling equipment.*
-
- -

It wasn't always this way. This is a story from early in the

technological revolution, when the application was out searching for the
hardware, from a time before the Internet, a time before the PC, before the
chip, before the m a i n h e . From a time even before programming itself.
'We find in 1990 "About 30 years have passed since computers were first
introduced into the humanities," Giacinta Spinosa, "Introduction," Computers and
the Humanities, 1990, p. 349. The 1960 terminus is essentially right, which makes it
the more important to look at-and appreciate-the work in the decade before it.
'A good example is Packard's development of the Ibycus from a Hewlett-Packard
computer.

The Classical Bulletin 75.1 (1999) 3-20

Let us step back by reviewing a series of key dates in computing, and
humanities computing. Some entries below are for products, some are for
tools.

World Wide Web. Tim Bemers-LeeICERN

Perseus Project planning starts.
Ibycus (mini-mainhe). David Packard.
Through 198GIndex Thornisticus: (hard-copy, 56
voll.) R BusaICAEL
TLG Planning Committee.
David Packard's program-produced Concordance
to Livy.
Thomas Aquinas text card-punching completed.
R. Busa4BM
(Published) Dead Sea Scrolls machine-readable.
R. BusaOBM
Magnetic-tape assisted Bible Concordance.
J. Ellison/Remington Rand.
FORTRAN made public.
Computers made public (IBM ships IBM 70 1)
Programming invented.
Machine-generated concordance. R. Busa/IBM
ENIAC (electronic tube computer).U.S.Army.
MARK I, IBM/Harvard, electronic relay computer.
U.S. Census recorded on punch-cards; Hollerith
founds IBM predecessor.
Punch-cards. Herman Hollerith patent.

To start with the 1951 entry is essentiallyto start with the beginning of
humanities computing. For fiuther setting and contrast, the U.S. Bureau of
the Census installed the fust UNIVAC in 1951, and as noted above, IBM
delivered its fmt computer, the 70 1, in 1953.3 With regard to the above
setting, how could you do a machine-generated concordance in 195l ?
To begin to answer, Tasman's 1957prophecy was no shot in the dark.
His view of the future was a projection from his recent past. Thomas J.
Watson, Sr. had assigned him in 1949 to be IBM liaison and support
person for a young Jesuit's daring project to produce an index to the
complete writings of St. Thomas Aquinas [Ref. 176, p. 841.
First, Tasman's thesis, as subsequent history turned out, was a huge
understatement, and second, it essentially defines the first large invention
of Father Roberto Busa, S.J., namely, to look at "tools developed primarily
for science and commerce" and to see other uses for them. As will be seen,
this was a case of fortune favoring the prepared mind. Redirecting

3Fisher, Franklin, et al. IBM and the US Data Processing Industry: an Economic
History, New York, 1983, p. 8 and 12.
 ROBERTO S.J.
BUSA,

ROBERTO
BUSA,
S.J.

business record-keeping equipment4 to the service of philological

scholarship, he essentially invented the machine-generated concordance,
the first of which he had published in 195 1.
Father Busa, of course, is best known as the producer of the landmark
56-volume Index Thornisti~us.~ As he began this work in 1946, and
produced a sample proof-of-concept, machine-generated concordance in

4The Computer-Tabulating-Recording Company (C-T-R),the conglomerate that

became IBM, initially marketed meat-cutters, cheese-cutters, and grocery-store
scales, in addition to registers and tabulators. See Rogers, William, THINK, a
Biography of the Watsons and ZBM, New York, 1969, p. 80.
1951, his professional life spans the entire computing chapter in the
history of scholarship. Emphasis in this article will be on the early steps.
He announced the need for such a project in his dissertation, and
made a presentation of the need for an Index Thomisticus at the
International Congress of Philosophy, Barcelona, October, 1948 [Ref. 3, p.
4251. The announcement of the project in Speculum [Ref. 3, p. 4251 starts
us at the handmade level of the Thesaum Linguae Latinae, and
envisioned a work in two parts:

a) A general file of all the words appearing in the works of St.

Thomas. Each card in the file will contain in the upper left hand
comer, a specific word used in the text of St. Thomas. Below that
will be given the exact reference to the place in his text where the
word appears, with a quotation of the sentence in which it is
found. Such a file would need about thirteen million cards.. .

b) Indices and Concordances to be drawn from such a file.

No hint of the revolution to come, except an appeal to scholars for

any information they can supply about such mechanical devices

as would serve to achieve the greatest possible accuracy, with a
maximum economy of human labor. (Father Busa has been in
contact with IBM in New York, the RCA laboratories in
Princeton, the Library of Congress, and the Library of the
Department of Agriculture, in Washington.)

Like all good projects, this one began with a question: What is the
metaphysics of presence in St. Thomas Aquinas? Combing for praesens
andpraesentia, he realized that such words were peripheral, and, however
unfortunately, Saint Thomas's doctrine of presence is linked with the
preposition in!
Inquiring what St. Thomas meant by "presence," the young Roberto
Busa realized that we must also study the way hction-words affect
meaning-words. To study the significant phrase "in the presence" he
needed the shades of "in." His dissertation, defended in 1946, was
essentially founded on a handmade Thomistic Concordance, essentially
complete, but with one entry.
He had made 10,000hand-written cards.
This project had results quite beyond the theological and
philosophicalvalue of his findings published in the first entry of the select
bibliography below. Deeming it necessary to learn what significance

SRoberto Busa, S.J.,Index Thomisticus: Sancti Thomas Operum Omnium Indices

et Concordantiae, Stuttgart, 56 voll, 1974-1980. This is now available, o f course, on
CD-ROM. References to other works of Fr. Busa are keyed to the select bibliography
that accompanies this article.
words have in an author's mind before attempting to gain insight into an
author's conceptual system, he envisioned a concordance of all the words
of St. Thomas Aquinas including the conjunctions, prepositions, and
pronouns, a vision which required dealing with 10,000,000 words: all
phrases broken out, each phrase copied over once for every word within
the phrase, the lemma indicated on each of these cards, then sorted....
Having dealt with "grand games of solitaire" with 10,000 cards, he knew
the scope. In short, he saw a need for the impossible.
Knowing this was impossible, he began a search for mechanical
assistance. In the United States, mechanical manipulation of cards had
already become a major industry. This industry was based on Herman
Hollerith's patents, and the Hollerith patents were owned and developed
by IBM.

Hollerith, a graduate engineer fkom Columbia University, had

developed a system of recording statistics for the U.S. Census
Bureau back in 1890 by punching holes in sheets of paper. By
adapting techniques he observed in the player piano and the
Jacquard weaving loom, which was regulated by cards in which
holes were punched to represent a pattern to be woven on the
loom, Hollerith had, in fact devised the forerunner to the
ubiquitous punch card with which IBM changed the world.6

Father Busa's survey of the mechanical record-keeping resources

took him to the United States in 1949. His coast-to-coast search took him
to about twenty-five American universities, a tour which led him to IBM
and to an audience with Thomas J. Watson.
Watson, known to business history as the founding genius of IBM,
was also a leader of the board of trustees of Columbia University, and had
recently been quite busy getting Columbia to recruit Dwight D.
Eisenhower as University President.' Given the requirements of the
project, and the card-machine resources under IBM patent, the meeting
was crucial, a make-or-break m ~ m e n t . ~

6Rogers,William, THNK, a biography of the Watsom and IBM, New York, 1969, p. 69.
'Rogers, pp. 203-208.
T h e situation of the then-contemporary data-processing industry has implications.
That Varia Specimina is the world's first machine-generated concordance approaches
the certainty of a theorem in geometry.
(a) Suppose the contrary, i.e., that there was a machine-generated concordance
before 1951.
(b) If it was done before 1951, it was done on punch-card manipulating machines.
(c) If it was done on punch-card manipulating machines of that era, it was produced
with the full cooperation of IBM.
(d) But this is the project done with that cooperation.
Item (b) above is a certainty. Item (c) above, is not fully as certain: Remington
Rand equipment is an outside possibility. The US Bureau of the Census lodged an anti-
trust suit against IBM before the war for monopolizing the punch-card
Father Busa recalled the meeting as follows:

I knew, the day I was to meet Thomas J. Watson, Sr., that he had
on his desk a report which said IBM machines could never do
what I wanted. I had seen in the waiting room a small poster
imprinted with the words:
"The diff~cultwe do right away; the impossible takes a
little longer."
(IBM always loved slogans.) I took it in with me into Mr.
Watson's office. Sitting down in front of him and sensing the
tremendous power of his mind, I was inspired to say: "It is not
right to say 'no' before you have tried." I took out the poster and
showed him his own slogan. He agreed that IBM would
cooperate..."provided that you do not change IBM into
International Busa Machines." [Ref. 176, p. 841

The first product of this alliance was formed partly by the limitations
of the then current IBM equipment. It would manage only eighty
characters on a card. Eighty characters? Father Busa realized this would
allow nothing longer than, say, a line of hendecasyllabic poetry. And a
lemma
So he did the poetry of St. Thomas Aquinas. His 1951machine-generated
and machine-printed concordance served as a proof-of-concept exercise.
How was it done?
In the preface to the Varia Specimina, Father Busa summarizes the
essential stages of work for generating concordances:

1. Transcription of text, broken down into phrases, on to

separate cards;
2 multiplication of the cards (as many as there are words on
each);
3. indicating on each of the resulting cards the respective entry
(lemma);
4. selection and alphabetization of all cards purely by spelling;
and
5. after the intelligentediting of the aphabetism,the typographical
composition of the pages for publishing.

data-processing industry. In that suit, it was revealed that IBM had a no-competition
agreement with Rernington Rand, whose card-machines used mechanical pins instead
of electric brushes to read the punch-outs. Remington in 1935 held the 15% of the
market left over from IBM (Rogers, 129-130). The question becomes, was there a
Remington Rand-associated concordance project? Not with punch-cards. The first
Remington Rand-associated concordance project appears to be John W. Ellison's
Bible Concordance o f 1957. It was done with the new magnetic tapes.
 ROBERTO
BUSA,
S.J.

At the time Father Busa entered the scene, only the second of these
concordance-tasks,the multiplicationof cards, was being done mechanically.
The TLL was using the services of a copying bureau; Professor Roy
Deferrari, Catholic University of Washington, was using electrical
typewriterswhich could make many copies; Professor P. O'Reilly of Notre
Dame University, South Bend, Indiana, had each side of the text-page
repeated as many times as there were words thereon. [Ref. 4, p. 221
Something was needed which could do all five. Promising was the
Rapid Selector, invented by Vannevar Bush and developed by Ralph
Shaw. Father Busa saw it operating in 1949 at the Library of the
Department of Agriculture in Washington, D. C. It simultaneously ran two
microfilm reels, one with text, the other with coded symbols for the words
of that text, and photographed the pages exhibiting the targeted
keywords. It examined 10,000 microfilmedpages per minute, instantaneously
rephotographing the found pages on another microfilm! [Ref. 4, p. 221
Though impressive, this looked like a blind end-all the coding had to be
done by hand! And it could not be adapted to automated printing.
One cannot but note that Father Busa knew the nature of the task and
knew what he was looking for. In the array of business machines in the
post-war IBM inventory, the punchcard electro-mechanical accounting
machines looked more promising than anything he had seen at the
American libraries or in other indexing projects.
What were these machines? How did they work? What was Father
Busa looking at-and later dealing with? No one has set this forth better
than Dr. Cuthbert Hurd, testifying in US vs IBM, the antitrust suit brought
by the Justice department in 1969.

The components of punched card equipment included brushes

which would detect...a hole in a punched card and which then
produced an electric signal, commutators which divided an electric
signal into a number of timing intervals, relays which opened and
closed..mechanical devices for punching holes in cards, and
mechanical printers. Relays could be opened and closed a few
dozen times a second and were subject to unreliable operation
because they were mechanicaland because of dust particles.. ..D M
had built avariety of machines using these components, including a
keypunch, ven3ersJinterpreter, reproducer,gangpunch, collator,
tabulator,sorter, and ~alculator.~

This is the array in which Father Busa managed to see the Index
Thornisticus. Computer? Not yet. Obviously applicable to Humanities
research? Not exactly. But to continue Hurd's testimony:

In Fisher, p. 13.
 These devices were controlled by control panels, or "plug
boards...." Such a control panel might measure three feet by two
feet and contain perhaps a thousand holes. Each machine had a
different control panel...manual intervention was the key and
because of manual intervention and because of the mechanical
nature of the devices, the results were slow and unreliable.
Consequently, there was a sharp limit on the size and kind of
application or tasks that could be performed.

Father Busa wrote in 1951, in the preface to his sample machine-

generated concordances:

What had at first appeared merely as intuition can today be

presented as accomplished fact: The punched card machines
cany out all the material part of the work in points 2,3,4, and 5.
p e f . 4, p. 221

Varia Specimina

Varia Specimina is a remarkable document for two essential reasons.

First, it was the demonstration that the punch card accounting machines
could do all five tasks essential to the production of a concordance.
Second, in describing the process in the preface, Father Busa gave us a
virtual time machine to see this process in its first generation.
First trials were done on one of Dante's Cantos. Each of the 136 canto
lines was punched (as position-coded holes) onto a card at the first
machine, the automatic punch. This was operated by a keyboard like a
typewriter, and was the only human input. Father Busa points out that an
error at this level would thus be perpetuated, but from this point on, error-
free would remain error-free:I0 no new ''typos" would have a chance to
creep in! [Ref. 4, p. 241
Proofreading was done by machine also, the Collator. Each line was
typed in twice. If the collator sees no difference, both are correct; if there is
a difference, a typographical error is flagged, and the line was re-punched.
The rest was largely a matter of overseeing the machines. At the next
station, the Record Interpreter machine printed what was 'written' in the
holes. At the next station, the Reproducer made a copy of each card, plus,
at the side, the first word of the line. Then a second copy, with at the side,
the second word of the line. Finally, there were as many cards as words in

"'This was an abiding theme for the entire project. In their write-ups of the time,
both Busa in Europe and Tasman in New York used as a facing card illustration
Thomas's words-reminiscent of Aristotle-QUIA PARWS ERROR IN PRINCIPIO
MAGNUS EST IN FINE.... It made a good sample punch card.
the text. In other words, each line was multiplied as many times as the
words it contained. At this point, alphabetizing was a matter of feeding the
Sorter Machine [Ref, 4, p. 261, and if a researcher were to drop the cards,
getting them in correct array would be (ideally) a matter of simply putting
them in the Sorter again.
Most of the manual intervention occurred here, as the business
machine did not begin the second word in the same column on all cards, or
the third, or the fourth.... The language itself was trouble-causing. E.g. in
the Italian o, ebbi, avrei, are all forms of the same word; andiamocene is
three words presented as one; in Latin rnortuus est is (functionally) one
word presented as two. When the necessary manual intervention was
done, the last stop was the Alphanumeric Accounting Machine, or
Tabulator. This retranscribed the words in the holes in the cards into the
letters and numbers. A print-out, in sum. As Father Busa put it in his
preface: "The concordance which I am presenting as an example is
precisely an off-set reproduction of tabulated sheets turned out by the
accounting machine." [Ref. 4,281

The prospectus fiom 1951 marked a turning point, and is worth

preserving:

"It is at all events certain that fiom now on the history of

concordances will no longer have to record figures like those of
the past: 500 Dominicans-can it really be true?--employed by
Hugh de St. Cher in 1200 in Paris for the first Biblical Latin
Concordance; fifty monks occupied in preparing the Biblical
concordances organized by the Benedictines in 1700; five
German Universities cooperating to set in order by hand the ten
million cards of the TLL at the end of the last century.

"Today it will suffice that man's hands transcribe by typing the

entire text on the Punch, control its accuracy either perusing the
cards which the Interpreter will have made legible or using, as I
have mentioned, the Verifier or the Collator; this done, the
Reproducer, the Sorter and the Accounting Machine and
Collator will take entire care of the remaining material part of the
work; in a few days the philologist, with the mere help of a
technical expert for the care of the machines, will have in hand
the general card file and the final proofs corrected for the printer,
certain of an accuracy which could never have been guaranteed
by the cooperation of man's sensorial and psychical nerve
centers." [Ref. 4, pp. 34,351

The 1951 demonstration that accounting machines could generate a

concordance offered six tools for scholars working with the poems of St.
Thomas:
 Words and their frequency, forwards
Words and their frequency, written backwards
Words set out under their lemmata (eg, aemulis under
aemulus aemula aemulum), with frequency
4. The Lemmata
5. Index of the words
6. Keyword in context Concordance

The project was done in IBM offices in Milan, where Father Busa
started his own punching and verifling department.

In 1954, he started a training school for keypunch operators.

"The success was excellent: industries wanted to hire them

before they had finished the program. Their training was in
punching and venJLing our texts." [Emphasis added.] The
school was continued until 1967, when the text-punching was
finished. He then moved his operations to Pisa, then in 1969 to
Boulder, Colorado, and in 1971 to Venice, "wherever IBM
provided computer time." [Ref. 176, p. 851

In the 1951project, no computers, no programming. Though there is no

attempt here to recount the history of the computer, it should be noted for
context that the fmt electronic (vacuum tube) computer was developed for
the U.S. Army and operational in 1946, the ENIAC. Prograrnming, the
storing of operating instructions in the machine so it could operate
independently, was invented in 1952, and the invention is usually credited
to John Von Neumann of the Atomic Energy Commission.' '
Back in the United States, Benjamin D. Wood, a pioneer in
educational measurement who had to oversee the scoring of 35,000
examinations a year, also got an appointment with Thomas J. Watson
senior. The immediate result was a convoy of trucks bearing card
machines to Columbia University-which was never billed for them. In
reviewing the work which these machines were doing, and interrupting a
gush of grateful superlatives, Watson asked Wood "What's wrong with
our machines?" He was not ready for the answer.

Wood replied they were too slow. "Ultimately these machines, which
are now electro-mechanical, since they operate on electricity, are going to
operate at the speed of light, 10,000times the speed at which they function
now." Watson's biographer reports

"THINK, p. 175.
 BUSA,S.J.
ROBERTO

"A gleam appeared in Watson's eyes...."I2

Only the tiniest fraction of the machine time per card was spent
reading its punches. Almost all was in the physical manipulation of the
card. The now-familiar mark-sense machine scoring was not possible.
Carbon from pencil marks had an electrical resistance, 500-5,000 ohms,
varying by a factor of ten. A Michigan school teacher, Reynold B.
Johnson, invented a scoring machine with such high resistance that the
pencilmark resistance was not consequential. IBM bought it.I3

The Transition of the Later Fifties

Programming in the Varia Specimina project had been a matter of

planning the process. The process was one of marking instructions on the
text or on a transparent overlay on the text for the keypunch operator, and
then overseeing the machines and overseeing the transport of cards from
machine to machine. When we next see a report of the hardware and
software behind the project, we see mark-sensing, a transition to magnetic
tape, and a mention of programming.
Learning in the press (apparently in 1954 [Ref. 25, p. 373n1) of John W.
Ellison's Bible Concordance work with Remington Rand magnetic
records, Father Busa then met him in person, shook his hand saying "You
are a great ally of mine," then immediately went back to IBM, where he
asked "See what Remington is doing?"Ref. 176. p. 851
Punch-card data-processing, though miraculous compared to hand-
work, was slow: Here is the view from 1958: "Mechanical alphabetizing
requires two passes of the cards through the machine for each column
sorted. Thus sorting 100,000 cards containing words of ten letters means
in effect, passing 2,000,000 cards through the machine. Depending on the
model sorter used, from 30,000 to 60,000 cards per hour can be sorted.
Therefore it could take from thirty-five to sixty-five hours, approximately,
to accomplish the alphabetization. In other words, the machine would
alphabetize h m 1,500 to 3,000 words of ten letten in an hour." p e f 25, p. 36 11
Even this, of course was appreciated. Paul Tasman produced this
comparison in 1957, when the 2000 pages of the Summa Theologica were
represented by 1,600,000 word-cards:

1. Manually, three persons, 20,000 hours.

2. Punch-card method: three persons, 1,000 hours.
3. "Large-scale data-processing method," one person working
60 h0urs.1~

"THINK, p. 138.
')THINK, p. 139.
The definition of what he meant by "Large-scale data-processing method"
glimmers in the next sentence: "This, of course, is exclusive of preparation
andprogrammingtime." [emphasis added] Programming is now in civilian
use, but it is still in card-or card-analog-management. l5
The cards are now, figuratively speaking, two-edged. Machine
punched for the one, mark-sense capable for the other. The same machine
can now read both holes and pencil-lead. That is, the cards enable built-in,
machine readable human intervention-for which, as we will see, there is
always great need.
Tasman gives the background to Father Busa's Dead Sea Scrolls project:

In this application, the major objective is compactness of

magnetic tape files and the speed at which tapes can be read,
written, and printed.

Each word card...is initally converted to magnetic tape in blocks

of records 80 characters long. The process is broken down into
four runs of the IBM 705. Run 1 deals with sorting and inverting
each word in memory. Run 2 deals with creating the frequency
count of the stored words and their summarization. Run 3 deals
with a merging and grouping of the different word tapes with the
entry word tape. Run 4 provides for collating each word tape
with the original phrase tape.16

Much of the actual "programming" was still in the design of the card, and
the essential over-all program was still visibly the process for producing a
concordance laid out by Father Busa in Varia Specimina. I venture to
summarize the 1957 algorithm, from Tasman's section "Specific phases of
automation in the literary analysis of the Summa Theologica of St.
Thomas Aquinas":

1. The scholar analyzes the text, marking it with precise

instructions for card punching, and phrase limits.
2 A clerk [a student in the keypunch operators school] copies
the text. The phrase is now holes in a card, preceded by a
reference, and a serial number. A second clerk retypes the
phrase on the same card [apparently a change from 19511.
Cards with discrepant phrases are found by a checking
machine, and replaced.

14Paul Tasman, July, 1957 IBM Journal of Research and Development, p. 256.
"The first FORTRAN system was released in 1957, for the IBM 704. This was the
year of Tasman's report/prospectus. FORTRAN was, of course, for numerical
computation. See S. Rarnsden, University of Manchester, http://www.man.ac.uk
hpctec/courses/Fortran9O/Fortran90~4.htrn1
I6Tasman, p. 256.
 ROBERTO
BUSA,S.J. 15

3. From the phrase-cards the machine [the IBM 705?] does two
jobs: a) produces the word cards, and b) produces a complete
copy of the text, phrase by phrase. In various zones of the
card, there are encoded 1)the reference, 2) the first letter of the
preceding word and the first letter of the following word, 3) the
ordinal number denoting the word's position in the text, 4) a
characterizing mark. l7
4. Form cards: the machine counts and eliminates all word
duplicates. In this stack, there is left only one card for every
graphically different word, and a record of its total
occurrences. At this stage, sum and est, for example, are still
. different entries.
5. The scholar converts the form-cards into the entry cards.

At this point in the algorithm, the scholar's task is two-fold, and

Herculean. He must first examine the form cards and group all the
different forms of the same semantic unit under the one word
which will serve as the entry, or lemma. He must then split all
homographs into their respective lemmas. For example, in
English, lead the metal must be separated from lead the verb; in
Latin, amor the noun from amor the passive first person singular
verb form, or labor the noun "work" from labor the deponent
verb "I am slipping."

This is morphology on the practical level. It is intense work. Professor

Busa has written a book about it which will reward anyone interested in
Latin. It is Inquisitiones Lexicologicae. [Ref. 2581

6. Where not already done, all groups of cards are "interpreted."

This means the machine prints on top of each card whatever
information is on or in the card.
7. "Information on any desired groups of these cards can now be
printed on sheets, in brochures or books or on other cards.
Valuable tools in philological research like an index verborum
or concordanceare available without further scholarly effort."

The work had interesting consequences for text-restoration. In Dead Sea

Scroll experiments, "up to five consecutive words have been 're-written'
by the data-processing machine in experimental tests where the words
were intentionally left out of the text and blank spots indicated."I8

''The characterizing marks? Some examples:

1 marked a passage where St. Thomas refers to other passages of his own work;
# marked the words of another author quoted by St. Thomas;
marked phrases not entirely Thomistic as St. Thomas recorded, discussed, or
refuted the doctrine of others.
18Tasman, p. 256n.
 Conclusion

More generally, Tasman foresaw a new era coming from the

techniques developed in the BusaflBM project:

The indexing and coding techniques developed by this method

offer a comparatively fast method of literature searching, and it
appears that the machine searching application may initiate a
new era of language engineering. [emphasis added] It should
certainly lead to improved and more sophisticated techniques
for use in libraries, chemical documents, and abstract preparation,
as well as in literary analysis.19

In conclusion, Father Busa, with IBM's enabling help, was at the

pivot point (or was the pivot point) between handmade scholarly tools
and machine-made scholarly tools. His impossible dream of 1949 is the
reality of 1999. We are in Tasman7s"new era of language engineering." As
for the Index Thornisticus, the goal which motivated the transition, it has
been completed and has undergone transformations from punch-cards to
tapes to books to CDs. There is a growing bibliography of theological and
linguistic work founded upon it.20

Appendix 1: A Challenge for FurtherWork

But, in the words of Father Busa himself in 1990, "Our generation has
not done everything: for the young people there are still immense open
spaces" [Ref. 243, p. 3 3 91. He posed eight challengesldesiderata:

1. The automatic linking of syntagmata-sets of more than one

word that should be treated as a unit, e.g. "surnames and
names, the compound form of verbs, expressions such as
'clothes horse. "'

2. The automatic identification of the word to which a pronoun

refers.

3. The detection of tacit words2'

"Tasman, p. 256.
20Somemay be sampled in Paolo Guietti, "Hermeneutic of Aquinas's Texts: Notes
on the Index Thornisticus," The Thomist, 57, (1993) pp. 667-686, where one may
also see a knowledgeable view of the Index.
"1 take "tacit words" to mean words which are sufficiently implied, but not present
as, for instance, in the Greek geometric texts, the word for "line" can be represented
by simply the feminine article or feminine adjective; "point" by the neuter.
 4. The creation of a general conspectus-within the confines of
a precise universe--of all words which are potentially
homographic.

5. For every homographic word, or at least for each type, the

recognition of what in the context is specific to one or the
other of its values.

6. The recognition of all elementary and direct grammatical, i.e.,

semantic connections. For instance, which words in a given
sentence are used as adjectives, and to which nouns do they
refer? what is the object of a given verb?

7. The definition of the logical function of every word or clause;

whether a subject, an object, a verb, or a complement.

8. Finally, the formalization of a logical train of thought winding

through many paragraphs; this would permit automatic
indexing and automatic ab~tracting.~~

Appendix 2: Select Bibliography

The publications of Father Busa outside of the Index Thomisticus

form a corpus of more than 200 entries, of which the following is a
significant selection.

1. La Teminologia Tomistica dell' Interiorita. Saggi di metodo per

una interpretazione della metafisica della presenza, Milano,
Bocca, 1949,pp. 280.

3. "Complete Index Verborurn of St. Thomas Aquinas," in Speculum-A

Journal ofMediaeva1Studies, X X V , (Jan. 1950),pp. 424425.

4. Sancti Thomae Aquinatis Hymnorum Ritualium Varia Specimina

Concordantiarum. Primo saggio di indici di parole automatica-
mente composti e stampati da macchine IBM a schede perforate,
Milano, Bocca, 1951,pp. 180.

"I have shortened or paraphrased some of these desiderata. They are in Ref. 243,
p. 341. Father Busa credits Peter Luhn of IBM with formulating challenge number 8
in the early 1950s, and also credits him as the introducer of the KWIC Index. Finally,
a personal note: working towards desideratum number five has, coincidentally, been
the core of my career as a researching Latinist since 1980-which is the vintage of
the IBM 360 mainframe in the background of Father Busa's picture.
7. "Mechanisierung der philologischen Analyse," in Nachrichtenfir
Dokumentation, vol. 3, n. 1 (Mar. 1952),pp. 14-19.

12. "La Cibernetica," in Ingegneria Meccanica, Milano, 111-2 (Feb. 1954),

pp. 5-10.

25. "The Use of Punched Cards in Linguistic Analysis," in Punched

Cards-Their Application to Science and Industry, 2nd edition,
R.S. Casey et al. ed., New York, Reinhold 1958,pp. 357-3 73.

28. "All non-biblical Dead Sea Scrolls published up to December 1957

have been indexed," in Sacra Pagina, Miscellanea Biblica
Congressus Internationalis de Re Biblica (1958) Paris, Lecoffre
1959,pp. 7-12.

30. "Erlauterungen zu den lexicographischen Arbeiten zu Goethe,"

Farbenlehre, Bd. 3 Kolloquium: Maschinelle Methoden der
literarischen Analyse und der Lexicographie. Tuebingen, 24/26 Nov.
1960, pp. 36 (litografia).

35. Atti del Convegno: Linguistica e Industria Oggi a cum di R. Busa

S.J. Milano, 19 Genn. 1962,pp. 75 (ciclostilato).

40. L' Analisi linguistics nell' evoluzione mondiale dei mezzi di

informazione in Almanacco Letterario Bompiani 1962, Milano,
1962,pp. 103-107.

4 1. Stellungnahme in "Zur mechanischen Sprachuebersetzung.J. J.

Becher, Allgemeine Verschlusselungder Sprachen" von Prof. W. G.
V. Waffenschmidt, Kohlammer Verlag Stuttgart, 1962,pp. 45-52.

45. "An inventory of fifteen million words," in Literary Data Processing

Conference Proceedings, 9/10/11 Sept. 1964, Modem Language
Assoc. New York, 1965, pp. 64-78.

47. "World Evolution in Information Processing and its Influence on

Linguistic Research" (in Hebrew), in Third World Congress of
Jewish Studies, Jerusalem 1961,Jerusalem, 1 965, pp. 4749.

48. Die Elektronentechnik in der Mechanisierung der sprachwissens-

schaftlichen Analyse (Traduzione in msso dell' articolo del1957) in
Avtomatizatsiva v lingvistike,L. N . Zasorina, Nauka, Moscow, 1966.

53. "Saggio esplorativo di automatic abstracting," in Convegno

Nazionale T.D. 66: La riduzione Concettuale dei Documenti.
Torino 14.12.1966. Atti a curadel CSAO Torino, 1967,pp. 3947.
 ROBERTO
BUSA,
S.J. 19

63. "Erreurs Humaines dans la Preparation de 1'Input pour Ordinateurs,"

in Les Machines dans la Linguistique. Colloque International sur
la mecanisation et l'automation des recherches linguistiques,
Accademia, edition de 1'Academie Tchecoslovaquedes Sciences,
Prague, 1968,pp. 279-284.

66. Actes du Seminaire International sur le Dictionnaire Latin de

Machine Pisa 27/28/29 Mar. 1968, redige par Roberto Busa, S.J.,
pp. 176. (Appeared as supplement n. 2 to vol. V of Calcolo, Inac, P.
le delle Scienze 7, Roma) 1968.

69. "L' Instrumentation electronique dans les recherches linguistiques,"

in Rencontre International de Mechanographie et Informatique,
Lisbonne Oct. 1967, Caixa Nacional de Pensdes, 1970,pp. 737-747.

7 1. "Concordances," in Encyclopedia of Library and Information

Science, Marcel Dekker, New York, vol. V 1971,pp. 592404.

73. "The Impact of Cybernetics on the Humanities," in Proceedings of the

Jurema 1972: International Symposium on Cybernetics in Modern
Science and Society, ed. W. Muljevic Zagreb, 1972,pp. 13-24.

104. "Inner and Outer Information: Causality types in man-to-man

information," in Proceedings of the Jurema 1975 International
Symposium on Cybernetics in Modern Science and Society, ed. W .
Muljevic, Zagreb, 1975,pp. 17-20.

1071108. Concordantiae Senecanae, R. Busa S.J., A. Zampolli, Georg

Olrns, Hildesheim, 1975.2voll., pp. IX+VII+1473+5!%59+5R
[Father Busa's masterful Seneca concordance is probably the work by which
he is known to classicists. Perhaps the best measure of its worth is that no one
has ever tried to supersede it; one thinks of the competing, and Jawed,
concordances of e.g. Vergil and Manilius. In sum, its trustworthiness and ease
of use have ensured that it remains a welcome vade mecum even in the age of
CD-ROMs. (Ed.)]

136. "Computer Processing of Over Ten Million Words: Retrospective

Criticism," in The Computer in Literary and Linguistic Studies:
Proceedings of the Third International Symposium, Alan Jones, R.
F. Churchouse, edd. The University of Wales Press, Cardiff, 1976,
pp. 114-1 17.

147. "Man-machine relationship in computerized linguistics," in XWII

Convegno Internazionale delle Comunicazioni, Genova 9/12 Ott.
1979,Istituto Internazionale delle Comunicazioni Genova, 1979,pp.
279-286.
149. "ORDO dans les oeuvres de St. Thomas d7Aquin," in ORDO 11
Coll. Internazionale Lessico Intellettuale Europeo, Roma 7/9
Genn. 1977,Ed. Ateneo-Bizzarri, Roma, 1979,pp. 59-1 84.

176. "The Annals of Humanities Computing: the Index Thomisticus,"

Computers and the Humanities, New York, 1980, pp. 83-90.

178. "Per S. Tommaso 'ratio seminalis' significa 'codice genetico' :

problemi e metodi di lessicologia e lessicografia tomisticha, in Atti
dell' VIII Congresso Tomistico Internazionale, vol. 1, Citta del
Vaticano, 1981,pp. 437-45 1.

193. Global Linguistic Statistical Methods to locate style identities:

Proceedings, ed. by R. Busa, S.J. Lessico Intellettuale Europeo,
Ateneo, Roma, 1982, pp. 111.

195. "Trente ans d' analyse informatique de textes: ou en est-on? et

apres?" in Actes du Congres International Informatique et
Sciences Humaines, Li2ge 18/21 Nov. 1981,Ed L.A.SoL.A. Liege,
1983,pp. 135-148.

200. "De texminationum Latinarum statisticismensuris ex Indice

Thomistico," in Hommage a Pierre Guiraud,Les Belles Lettres,
Nice, 1985,pp. 340.

206. "Informatica e Nuova Filologia," in Lessicolografia, Filologia e

Critica--Convegno Internazionale di Studi Catania-Siracusa 26/28
Apr. 1985Atti, Ed. Olschki Firenze, 1986,pp. 200.

208. "De Linguae Latinae flexivis terminationibus," Revue Informatique

et Statistique dans les Sciences Humaines, XXI 1-4, Liege, 1985,
pp. 53-66.

215. '2' originalitelinguistique de St. Thomas d'Aquin," ed. A.L.M.A.

XLN-XLV, Brill-Leiden, 1985,pp. 66-90.

2 19. Fondamenti di Informatica Linguistics, Ed. Vita e Pensiero,

Milano, 1987,pp. 4 12.

220. "Das Problem der Thomistichen Hermeneutik nach der Veroffent-

lichung des Index Thomisticus," Miscellanea Medievalia, vol. 19,
Koln, 1988,pp. 359-364

222. "De phantasia et imaginatione iuxta S. Thomarn," in Phantasia-

Imaginatio: V Coll. Internazionale Lessico Intellettuale Europeo-
Roma 9/11 Gen. 1986Atti. Ateneo Roma, 1988, pp. 135-1 52.
225. "Inteligencia natural e inteligencia artificial," Broteria, Lisbon, 1988,
pp. 260-268.

229. Totius Latinitatis Lemmata - quae ex Aeg. Forcellinii Patavina

Edizione 1940 a fionte, a tergo atque morphologice, opera IBM
automati ordinaverat Robertus Busa S.J., Istituto Lombardo di
Scienzee Lettere Milano, 1988,pp. XVI + 532.

230. "Procedures et resultats de la segmentation thematique des lemmes

latins de 1'Index Thornisticus," Revue Informatique et Statistique
darts les Sciences Humaines, 24, 1-4 Liege, 1988.

243. "Informatics and New Philology," Computers and the Humanities,

New York, 1990,pp. 339-343.

244. "Idea negli scritti di Tornrnaso D' Aquino," in Idea - VI Coll.

Internazionale Lessico Intellettuale Europeo - Roma 5/7 Genn.
1989Atti, Ateneo Roma, 1991, pp. 63-87.

248. "De Expressione apud S. Thornam," in Littera Sensus Sententia -

Studi in onore di C. L. Van Steenkiste, OP,Ed. Massimo Milano,
1991,pp. 135-154.

250. "Cinguant' anni a 'bitizzar' parole," in Convegno Internazionale

sulla Storia e Preistoria del Calcolo Automatico e dell'
Informatics - Siena 10/12 Sett. 1991 Atti Precongressuali, Siena,
1991,pp. 72-82.

25 1. "Half a Century of literary Computing: Towards a 'new' Philology,"

in International Conference on Current Issues in Computational
Linguistics-Penang 11/14 June 1991 Atti Precongressuali,
Penang, 1991, pp. 84-95.

257. Thomae Aquinatis Opera Omnia (cum hypertextibus) in CD ROM

Milano Editel 1991,l CD ROM + 64 pp.

258. Inquisitiones Lexicologicae in Indicem Thornisticum, Milan, with

accompanyingEnglish tr.by Philip Barras, Milan, 1994.

THOMAS NELSON WINTER

University of Nebraska