What, if anything, is typology?

JOHANNA NICHOLS

Typology has the hallmarks of a mature discipline: a society, conferences,

journals, books, textbooks, classic works, a founding father, and people who
are called and call themselves typologists. A typologist probably teaches a
course with a title like “Typology and Universals” which includes readings by
Greenberg, Dixon, and Dryer, often a textbook such as Whaley (1997), Comrie
(1989), Song (2001), and/or Croft (2003), and some grammar-reading assign-
ments. With regard to research, the typologist reads grammars, does at least
some crosslinguistic research, has probably done some fieldwork and descrip-
tion, and usually does not identify with or claim allegiance to any particular
named theoretical framework.1 Despite these conspicuous identifying marks, I
submit that the position of typologists on this should be that there is no such
subfield of linguistics as the usual referent of “typology”.
As the essentials of typology, I would offer the following: framework-neutral
definitions (see again Footnote 1), emphasis on codability in definitions and in
applications of theory, bottom-up or data-driven constructs, and concern with
observable phenomena that pattern interestingly in the world’s languages (in
their frequency, their interaction with other parts of grammar, their geography,
their history, etc.). Theoretical constructs are built on a very simple machin-
ery: a methodology of survey design common to most social and biological
sciences, and classic structuralist notions such as contrast and complementary
distribution. Over the last four decades, typological theory has built up a body
of knowledge focusing on the building blocks of grammar and their distribu-
tions, and another body of knowledge having to do with survey methods, sam-
pling, significance testing, defining families and geographical areas, etc. Ex-

1. Henceforth, I use the term “framework” to refer to the formal systems that are usually referred
to as “theories” in formal grammar. A framework typically has a name, some of its own
terminology and formal notation, and its way of representing grammatical relations (e.g.,
trees).

Linguistic Typology 11 (2007), 231–238 1430–0532/2007/011-0231

DOI 10.1515/LINGTY.2007.017 ©Walter de Gruyter

Brought to you by | University of California - Berkeley

Authenticated
Download Date | 9/19/17 10:22 AM
232 Johanna Nichols

planations of these patterns focus on such things as how they arise, how they
function in language use, and how complex or straightforward their connec-
tion to the rest of the grammar is. Typological knowledge is mostly statistical
and probabilistic, not categorical. Typological theory is much like what Dixon
(1997: 128–135) calls “Basic Linguistic Theory” (though I would remove the
capital letters because it is not a framework but rather a framework-neutral the-
ory): the body of knowledge about grammar built up over the years by analysis
and comparison of different languages (my definition). Because typology seeks
durable and non-hermetic knowledge, typological theory is the same body of
knowledge as what field grammarians contribute to and draw on. For the same
reason, typology has a successful record of building up a cumulative and acces-
sible body of knowledge covering the entire history of the field, an area where
formal grammar and especially formal syntax have been less successful.2,3
Typology is perhaps more revealingly characterized not by what it is or does,
but by what it isn’t or doesn’t do. Despite the fact that typology courses and
textbooks usually feature the word “universals”, there is nothing like UG in ty-
pology. Though many typologists are interested in psycholinguistics and cog-
nitive science (and vice versa), knowledge of language is not an issue in ty-
pological theory; nor is innateness, despite typologists’ attention to universals.
Typological theory is almost entirely unconcerned with distinguishing possi-
ble from impossible languages, though typological distributional analysis is
sometimes concerned with distinguishing rarity from impossibility. The con-
cern in formal grammar with designing frameworks, theories, notations, and
other abstractions so as to preclude impossible languages has no counterpart in
typology. We have a curious split here: One part of linguistics (formal gram-
mar) pursues abstractions that sanction the universal and the possible, while
a different part (typology) does the concrete crosslinguistic work that might
empirically identify universals, and there is probably less communication here
than between any other activities of the two fields.4
Most of the work that typologists actually do is developing, refining, and ap-
plying typological theory. Most journal publications and conference papers in
typology are of this sort, often presenting a phenomenon from one or a few lan-
guages and laying out its implications for theory. With this assertion as back-
ground, let me address four misunderstandings about typology that seem to be

2. I believe this explains why, after 40 years of reading linguistics extensively, I find it easier to
read and understand the impact of a random article in an unfamiliar biological, earth, or social
science than a journal article in formal syntax dealing with a language that I know.
3. Dryer (2006) explicitly calls formal frameworks such as minimalism and optimality theory
“transient theories”.
4. Probably this is because, as Dryer (2006) observes, in typology and in field linguistics, ex-
planation of why language is as it is is kept distinct from description of languages, while in
formal grammar the explanation is also the description. See also Footnote 7.

Brought to you by | University of California - Berkeley

Authenticated
Download Date | 9/19/17 10:22 AM
 What, if anything, is typology? 233

widely believed outside of typology (and sometimes also by typologists them-

selves).
The first is that typology deals with only superficial grammatical phenom-
ena, while formal grammar deals with deeper abstractions (e.g., Newmeyer
1998, Baker & McCloskey 2007). It is true that until fifteen or twenty years
ago, anything larger than a small sample was necessarily restricted to what I
call “lookup characters” – the relatively superficial structural properties that
one could find in most of the then available grammars simply by looking them
up. Individual grammars, of course, contained in-depth analyses, but not all of
them on the same points, and most of the available grammars antedated the
advances made in the analysis of clause and inter-clause morphosyntax since
the late 1970s. By now, things are very different. I see no difference in an-
alytic or theoretical profundity or abstraction between generative parameters
and original contributions of typology such as direct object vs. primary object
(Dryer 1986), verb-framed vs. satellite-framed lexicalization patterns (Talmy
1985, Slobin 2004), various aspects of alignment (e.g., Dixon 1994, Dixon &
Aikhenvald (eds.) 2000), differential object marking (Bossong 1998, Aissen
2003), referential density (Bickel 2003), and others.
The second misunderstanding is that typology usually or often uses large
surveys of hundreds of languages. Most of the typological works quoted in
the paragraph above use samples of a few tens of languages, and Greenberg
(1963) famously used only 30. Large surveys are in fact not very common.
They are used when sampling and sample size itself is an issue (e.g., Dryer
1989), when the information is readily available and might as well be used
(as in most work on word order; published data appendices are intended to be
imported into other linguists’ databases, and subsequently the data points are
there anyway at no cost in terms of time or money), and in reference works
(e.g., Haspelmath et al. (eds.) 2005, for which sheer quantity of data on at
least some maps was an editorial goal that partly had the lay public as well as
reference and acquisition librarians in mind). Large samples are not pursued
for their own sake. In fact, samples of any kind are not frequent in typology
articles and conference presentations: most papers present an analysis based
mostly on a few languages; some draw examples from a variety of languages;
a quick survey of the last few years of Linguistic Typology indicates that only
two or three articles per year use formal samples.
It is rare for a study proposing a new typological metric or analysis to also
use a large sample; the exceptions I can think of are dissertations (e.g., Wälchli
2003/2005, Cysouw 2001/2003, Peterson 1999, Gensler 1993, and others) in
which a young scholar builds the foundation of a career by both pursuing
an analysis and creating the core of a database. Typological metrics are of-
ten first worked out on very small samples: the insight behind Talmy (1986)
grew out of comparing Atsugewi lexicalization with that of English; Bickel

Brought to you by | University of California - Berkeley

Authenticated
Download Date | 9/19/17 10:22 AM
234 Johanna Nichols

(2003) uses three languages; Gensler (1993) grew out of comparing two fam-
ilies. In my own work, the germ of head/dependent marking grew out of jux-
taposing Abkhaz and Ingush, was worked out on a convenience sample of a
dozen or so languages whose descriptions I knew well, was demonstrated on
a 60-language sample (Nichols 1986), and its geographical and genealogical
distribution was described with 174 languages (Nichols 1992). The transitiviz-
ing/detransitivizing lexical metric grew out of trying to make sense of Ingush
vs. Russian verbal lexicalization (Nichols 1982), the preliminary typology was
based on a twelve-language pilot study, and the metric was demonstrated with
an 80-language sample (Nichols et al. 2004).
Combining this observation with those of Baker & McCloskey (2007), it
seems that a new idea, either typological or formal, often emerges in in-depth
work on one or two languages (sometimes those are the field language and
the contact language, the ultimate in convenience samples). Hypothesis-raising
work, or working out the basic typology, is done on a small sample (Baker &
McCloskey’s “middle way”), and I believe most typologists rely for this pur-
pose on a small set of languages that they know well or whose descriptions they
know well. Hypothesis-raising work gives us most of our grammatical abstrac-
tions, whether typological or formal. Typological hypothesis-testing work of
various kinds uses larger samples and seeks, e.g., statistically significant corre-
lations between grammatical properties, between areas, etc., while the next step
for formal grammar is considering the impact on the wider framework. That is,
both are concerned with how the phenomenon fits into grammar; typology is
more explicitly crosslinguistic in this concern, but the main difference between
the two is framework-neutral vs. framework-oriented analyses.
The third misunderstanding is that in typology, explanations or theory are
usually functionalist. I find this claim (e.g., Baker & McCloskey 2007, Polin-
sky & Kluender 2007) startling because my own interest is pure structure and
its distribution. Explanation and hypotheses in typology come from all quar-
ters – function, processing, cognition, acquisition, neuroanatomy, sociolinguis-
tics, history, and language evolution (including the mathematics thereof, as in
Maslova 2000).
The fourth misunderstanding is that the main theoretical constructs of typol-
ogy are the implicational correlation and the implicational hierarchy. Implica-
tional statements are nicely testable in crosslinguistic surveys, so they are fairly
conspicuous in the research design for surveys, but they should be viewed as
a convenient format for presenting and testing results, rather than as the be-all
and end-all of typology.
In summary, most of what is done in typology is analysis and develop-
ment of typological theory. But much of what is said about typology con-
cerns implicational correlations, large samples, functional explanations, and
superficial characters. This mismatch is partly because what goes on in the

Brought to you by | University of California - Berkeley

Authenticated
Download Date | 9/19/17 10:22 AM
 What, if anything, is typology? 235

field called typology is two things: developing framework-neutral grammati-

cal theory (shared with field and descriptive work) and applying that theory
to crosslinguistic distributions and their implications. Perhaps only the latter
of these should properly be called typology, but, again, it is the former that
predominates in the journals and conferences.
Recent years have brought two reversals of a decades-long marginalization
of non-formal linguistics, both of which have strengthened the position specif-
ically of typology. Beginning in the late 1950s, field and descriptive linguistics
were not merely marginalized, but dropped entirely from some linguistics de-
partments and treated caustically in public discourse by formal theoreticians.
As a consequence, field linguists lost valuable time trying to construct their
own theoretical frameworks, students and faculty downplayed or abandoned
fieldwork, and many languages – including most indigenous North Ameri-
can languages – lost their last fluent speakers without ever receiving proper
syntactic description, adequate textual documentation, or revitalization efforts.
The stunning reversal of this trend since the appearance of Hale et al. (1992)
has brought the benefits of good linguistic description to the rest of the world
in time, brought prestige and funding to description and documentation, and
raised the profile of typology – rather than formal theory – as the source of
fieldwork questionnaires and the standard of sophistication in field description
and electronic documentation (this despite the long-standing computational ex-
pertise of formal linguistics).5,6 Typologically, rather than formally, oriented
linguists are the more successful recipients of grants for documentation of en-
dangered languages. The reason for this success is certainly not demography –
during this time, most linguistics departments at least in North America have
contained zero to one typologists. Rather, it must be the framework-neutral def-
initions, the growing body of substantive knowledge, and statistical and proba-
bilistic knowledge, all of which are readily applicable to description, compar-

5. I would actually date the germ of this development to the late 1970s or early 1980s when the
school of combined typology and field-oriented linguistics catalyzed by Bob Dixon in Aus-
tralia began to have impact on theory, and its young scholars began participating widely and
non-divisively in both formal and typological linguistics. In the same time frame, relational
grammar began to feed naturally and easily into crosslinguistic work.
6. My own experience working on long-distance reflexivization in Chechen and Ingush some
years ago may be instructive. I wanted to determine what these languages could contribute to
formal theory and what I should cover in order to make a description useful to formal the-
ory, and I read every formal work on distant reflexivization that I could find. It was almost
impossible to find out which phenomena were of interest, what I should elicit, and whether
long-distance reflexivization in Chechen and Ingush confirmed or falsified formal theoretical
analyses. In contrast, the bottom-up and at least typology-friendly analysis of clause com-
bining in Foley & Van Valin (1984) made very clear what I needed to elicit and what was
interesting.

Brought to you by | University of California - Berkeley

Authenticated
Download Date | 9/19/17 10:22 AM
236 Johanna Nichols

ison, and pinpointing what is distinctive and valuable about a particular lan-
guage.
The same half-century saw a similar (if less caustic) marginalization of his-
torical linguistics, which survived in language departments (though in recent
years, even these positions have often been administratively cut or converted
to pedagogical supervision) but lost most of its visibility and proponents in
mainstream linguistics. A consequence of the reduced staffing and prestige of
historical linguistics was the rough-roading and stultification of historical lin-
guistic knowledge by geneticists, archaeologists, and other prehistorians that
occurred the late twentieth century (e.g., Renfrew 1987, Cavalli-Sforza et al.
1988; the trend is lessening but still continuing today). Here, too, it is typology
that has benefited from and contributed to efforts to reach deep into linguistic
prehistory (Dunn et al. 2005, Bickel & Nichols 2005, Bickel 2003, Maslova
2000, Nichols & Peterson 1996, Nichols 1992, also Bickel 2007). The reasons
appear to be that codable framework-neutral definitions are needed in computa-
tional phylogeny, that rigorous crosslinguistic comparison can help determine
which structural features of language are most resistant to change and borrow-
ing, that geographical breakdowns in crosslinguistic surveys prove very useful
in tracing prehistory, and that strict comparison is basic for reconstruction.
In short, I suggest that what we call typology is not properly a subfield of lin-
guistics but is simply framework-neutral analysis and theory plus some of the
common applications of such analysis (which include crosslinguistic compari-
son, geographical mapping, cladistics, and reconstruction). Whatever typology
is, it is on a roll at the moment and likely to continue. By now, descriptive cov-
erage of languages worldwide, computational tools and expertise, genealogical
classification, and understanding of research design are adequate to support
comparison based not only on lookup characters, but on more complex and ab-
stract characters. If one of the blots on the record of linguistics is the inaction of
its mainstream in the face of language extinction in the second half of the twen-
tieth century, among its great accomplishments is the rapid expansion – both
qualitative and quantitative – of description, documentation, and classification
of the world’s remaining languages that is flourishing today.7

Received: 17 December 2005 University of California at Berkeley

Revised: 30 December 2006

7. As this paper was going to press, Dryer (2006) was published, a sophisticated analysis of
theories which notes (before this paper) that Dixon’s basic linguistic theory is the theoretical
bedrock of field and descriptive linguistics and much of typology, and observes that today’s
linguists confuse explanation with grammatical theory and therefore describe basic linguistic
theory as atheoretical, eclectic, etc.

Brought to you by | University of California - Berkeley

Authenticated
Download Date | 9/19/17 10:22 AM
 What, if anything, is typology? 237

Correspondence address: Slavic Department, University of California at Berkeley, Berkeley, CA

94720-2979, U.S.A.; e-mail: [email protected]

References
Aissen, Judith (2003). Differential object marking: Iconicity vs. economy. Natural Language &
Linguistic Theory 21: 435–483.
Baker, Mark C. & Jim McCloskey (2007). On the relationship of typology to theoretical syntax.
Linguistic Typology 11: 285–296.
Bickel, Balthasar (2003). Referential density in discourse and syntactic typology. Language 79:
708–736.
— (2007). Typology in the 21st century: Major current developments. Linguistic Typology 11:
239–251.
Bickel, Balthasar & Johanna Nichols (2005). Inclusive/exclusive as person vs. number categories
worldwide. In Elena Filimonova (ed.), Clusivity, 47–70. Amsterdam: Benjamins.
Bossong, Georg (1998). Le marquage différential de l’objet dans les langues d’Europe. In Jack
Feuillet (ed.), Actance et valence dans les langues de l’Europe, 193–258. Berlin: Mouton de
Gruyter.
Cavalli-Sforza, Luigi Luca, Alberto Piazza, Paolo Menozzi, & Joanna Mountain (1988). Recon-
struction of human evolution: Bringing together genetic, archaeological, and linguistic data.
Proceedings of the National Academy of Science of the U.S.A. 85: 6002–6006.
Comrie, Bernard (1989). Language Universals and Linguistic Typology. 2nd edition. Chicago:
University of Chicago Press.
Croft, William (2003). Typology and Universals. 2nd edition. Cambridge: Cambridge University
Press.
Cysouw, Michael (2003). The Paradigmatic Structure of Person Marking. Oxford: Oxford Univer-
sity Press. Based on doctoral dissertation, Katholieke Universiteit Nijmegen, 2001.
Dixon, R. M. W. (1994). Ergativity. Cambridge: Cambridge University Press.
— (1997). The Rise and Fall of Languages. Cambridge: Cambridge University Press.
Dixon, R. M. W. & Alexandra Y. Aikhenvald (eds.) (2000). Changing Valency: Case Studies in
Transitivity. Cambridge: Cambridge University Press.
Dryer, Matthew (1986). Primary objects, secondary objects, and antidative. Language 62: 808–
845.
— (1989). Large linguistic areas and language sampling. Studies in Language 13: 257–292.
— (2006). Descriptive theories, explanatory theories, and basic linguistic theory. In Felix Ameka,
Alan Dench, & Nicholas Evans (eds.), Catching Language: Issues in Grammar Writing.
Berlin: Mouton de Gruyter.
Dunn, Michael, Angela Terrill, Ger P. Reesink, Robert A. Foley, & Stephen C. Levinson (2005).
Structural phylogenetics and the reconstruction of ancient language history. Science 309:
2072–2075.
Foley, William A. & Robert D. Van Valin, Jr. (1984). Functional Syntax and Universal Grammar.
Cambridge: Cambridge University Press.
Gensler, Orin D. (1993). A typological evaluation of Celtic/Hamito-Semitic syntactic parallels.
Doctoral dissertation, University of California at Berkeley.
Greenberg, Joseph H. (1963). Some universals of language with particular reference to the or-
der of meaningful elements. In Joseph H. Greenberg (ed.), Universals of Language, 73–113.
Cambridge, Mass.: MIT Press.
Hale, Kenneth L., Michael Krauss, Lucille J. Watahomigie, Akira Y. Yamamoto, Colette Craig,
LaVerne Masayesva Jeanne, & Nora C. England (1992). Endangered languages. Language
68: 1–42.
Haspelmath, Martin, Matthew S. Dryer, David Gil, & Bernard Comrie (eds.) (2005). World Atlas
of Language Structures. Oxford: Oxford University Press.

Brought to you by | University of California - Berkeley

Authenticated
Download Date | 9/19/17 10:22 AM
238 Johanna Nichols

Maslova, Elena (2000). A dynamic approach to the verification of distributional universals. Lin-
guistic Typology 4: 307–333.
Newmeyer, Frederick J. (1998). The irrelevance of typology for grammatical theory. Syntaxis 1:
161–197.
Nichols, Johanna (1982). Ingush transitivization and detransitivization. Berkeley Linguistics Soci-
ety 8: 445–462.
— (1986). Head-marking and dependent-marking grammar. Language 66: 56–119.
— (1992). Linguistic Diversity in Space and Time. Chicago: University of Chicago Press.
Nichols, Johanna & David A. Peterson (1996). The Amerind personal pronouns. Language 72:
336–371.
Nichols, Johanna, David A. Peterson, & Jonathan Barnes (2004). Transitivizing and detransitiviz-
ing languages. Linguistic Typology 8: 149–211.
Peterson, David A. (1999). Discourse-functional, historical, and typological aspects of applicative
constructions. Doctoral dissertation, University of California at Berkeley.
Polinsky, Maria & Robert Kluender (2007). Linguistic typology and theory construction: Common
challenges ahead. Linguistic Typology 11: 273–283.
Renfrew, Colin (1987). Archaeology and Language: The Puzzle of Indo-European Origins. Cam-
bridge: Cambridge University Press.
Slobin, Dan I. (2004). The many ways to search for a frog: Linguistic typology and the expression
of motion events. In Sven Strömqvist & Ludo Verhoeven (eds.), Relating Events in Narrative,
Volume 2: Typological and Contextual Perspectives, 219–257. Mahwah, N.J.: Erlbaum.
Song, Jae Jung (2001). Linguistic Typology: Morphology and Syntax. Harlow: Longman.
Talmy, Leonard (1985). Lexicalization patterns: Semantic structure in lexical forms. In Timothy
Shopen (ed.), Language Typology and Lexical Description, Volume 3: Grammatical Cate-
gories and the Lexicon, 57–149. Cambridge: Cambridge University Press.
Wälchli, Bernhard (2005). Co-compounds and Natural Coordination. New York: Oxford Univer-
sity Press. Based on doctoral dissertation, Stockholms universitet, 2003.
Whaley, Lindsay J. (1997). Introduction to Typology: The Unity and Diversity of Language. Thou-
sand Oaks, Cal.: Sage.

Brought to you by | University of California - Berkeley

Authenticated
Download Date | 9/19/17 10:22 AM