ChapterPDF Available

Category genesis in Chitimacha: A constructional approach

Authors:
  • Chitimacha Tribe of Louisiana

Abstract and Figures

The genesis of new lexical categories poses a challenge to theories of diachronic change: If there are no pre-existing words in the class to analogize to, how does the category arise? This paper shows that a constructional approach to category change successfully accounts for the genesis of a diverse class of preverbs in Chitimacha, an isolate of the U.S. Southeast linguistic area. It is shown that what enabled the creation of the preverb category was schematization across a variety of forms with similar properties, namely, a preverbal syntactic position and a directional semantics. Category genesis can therefore be viewed as simply a special case of constructionalization wherein schematization plays a crucial role.
Content may be subject to copyright.
is is a contribution from Category Change from a Constructional Perspective.
Edited by Kristel Van Goethem, Muriel Norde, Evie Coussé and Gudrun Vanderbauwhede.
© 2018. John Benjamins Publishing Company
is electronic le may not be altered in any way.
e author(s) of this article is/are permitted to use this PDF le to generate printed copies to
be used by way of oprints, for their personal use only.
Permission is granted by the publishers to post this le on a closed server which is accessible
to members (students and sta) only of the author’s/s’ institute, it is not permitted to post
this PDF on the open internet.
For any other use of this material prior written permission should be obtained from the
publishers or through the Copyright Clearance Center (for USA: www.copyright.com).
Please contact rights@benjamins.nl or consult our website: www.benjamins.com
Tables of Contents, abstracts and guidelines are available at www.benjamins.com
John Benjamins Publishing Company

Category genesis in Chitimacha
A constructional approach
Daniel W. Hieber
University of California, Santa Barbara
e genesis of new lexical categories poses a challenge to theories of diachronic
change: If there are no pre-existing words in the class to analogize to, how does
the category arise? is paper shows that a constructional approach to category
change successfully accounts for the genesis of a diverse class of preverbs in
Chitimacha, an isolate of the U.S. Southeast linguistic area. It is shown that what
enabled the creation of the preverb category was schematization across a variety
of forms with similar properties, namely, a preverbal syntactic position and a di-
rectional semantics. Category genesis can therefore be viewed as simply a special
case of constructionalization wherein schematization plays a crucial role.
Keywords: Chitimacha, category genesis, schematicity, schematization,
constructionalization, preverbs
. Introduction
Category genesis presents a potential problem for theories of diachronic change that
rely on analogy as a key mechanism. When diachronic changes result in the creation
of an entirely new word class, there are no pre-existing words on which an analogy
could have been based. How then does the category arise? A construction-based
theory of diachronic change oers a solution in that it recognizes the existence of
schematicity, or abstractions across sets of constructions (Traugott & Trousdale,
2013, p. 14; Tuggy, 2007). Using data from the Chitimacha language, an isolate of
the U.S. Southeast linguistic area, this paper shows that a series of micro-level con-
structional changes (i.e., changes which aect the internal features of a construction
without creating a new one (Traugott & Trousdale, 2013, p. 1)), combined with an
increase in schematicity across otherwise unrelated constructions, can lead to the
subsequent creation of a new category in the language. It adds to the burgeoning
literature on reconstruction from a constructional perspective (cf. Barðdal etal.,
 ./cal..hie
©  John Benjamins Publishing Company
© 2018. John Benjamins Publishing Company
All rights reserved
 Daniel W. Hieber
2015), by applying diachronic construction grammar to the internal reconstruction
of categories within a language.
Chitimacha has a small class of preverbs whose members appear to have fol-
lowed dierent diachronic pathways and have origins in dierent word classes,
and yet were all reanalyzed as members of the same, new category of preverbs.
Since this was a new class of words, its emergence could not have been based on
analogy to already-existing preverbs, at least not for its rst members. Instead,
as will be shown here, the genesis of this category likely arose from analogy be-
tween similar constructions that all shared certain properties. is sort of ‘light
paradigmaticity’
1 among unrelated forms (i.e., schematicity) allowed for a parallel
and mutually-reinforcing process of constructionalization, i.e., “the formation of
new units (constructions) out of hitherto independent material” (Bergs & Diewald,
2008, p. 4), giving rise to the category of preverb. ese disparate words underwent
reanalysis to belong to the same, new word class on the basis of their common
constructional properties. is process by which constructions are reanalyzed to
conform to a newly recognized schema is what I term schematization.
is paper proceeds as follows: First I provide background on the language and
its system of preverbs. Next, I describe each of the nine preverbs and the evidence
for their diachronic origins. Finally, I sketch the process by which these preverbs
of disparate origins could have converged into a single category through a process
of constructionalization, and then conclude.
. Background
Chitimacha is a linguistic isolate spoken along the coast of Louisiana from the
time of French contact in 1699 until the last uent speaker passed away in 1940.
From 1930–1934, then-graduate student Morris Swadesh visited the Chitimacha
reservation in Charenton, Louisiana, and lled 16 composition notebooks with
texts and elicited sentences. Based on these materials, he prepared dra versions
of a grammar, dictionary, and text collection for the language, but these were never
published, and today these manuscripts are curated at the American Philosophical
Society Library in Philadelphia, PA (Swadesh, 1939a). It is Swadesh’s unpublished
but nearly-nished text collection (1939b) which constitute the data for the present
study, provided courtesy of the Chitimacha Tribe and the American Philosophical
Society Library.
. anks to Marianne Mithun for suggesting this useful term.
© 2018. John Benjamins Publishing Company
All rights reserved
Chapter2. Category genesis in Chitimacha 
e resulting corpus consists of 88 texts by one speaker, Benjamin Paul (chief
of the tribe from 1903 until his death in 1934), comprised mainly of traditional
narratives, but also some procedural and expository texts. Texts from the second
speaker, Benjamin Paul’s niece Delphine Ducloux, have not yet been digitally tran-
scribed, and so could not be included in this study. e available corpus contains
29,028 tokens of 4,467 types. e number of lemmas is currently unknown, but an
estimate from Swadesh’s dictionary manuscript (1939c), plus my own in-progress
database, would be approximately 3,700 for the entirety of the collection.
It must be noted that, because Chitimacha is an isolate, the reconstructions of
the etymologies provided here are inferred on the basis of internal evidence, most
prominently morphosyntactic reconstruction. e synchronic behavior of the pre-
verbs, on the other hand, is well documented in the corpus. While many linguists
see internal reconstruction as less robust than comparative reconstruction, years
of working with the Chitimacha corpus has taught me just how rich the insights
from internal reconstruction can be. us I am inclined to agree with Givón (2000)
that internal reconstruction is, properly applied, a sound and fertile method for
understanding language history.
. Preverbs in Chitimacha
A preverb is denitionally a category in ux. It is sometimes characterized as a
‘separable verb prex’, sometimes as a cover term for preverbal words and preverbal
prexes (Los etal., 2012). is is because a common feature of all denitions of
preverbs is variability in their syntactic freedom, where certain preverbs are more
tightly bound syntactically to the verb and have more functional meaning, and
other preverbs may be syntactically separated from the verb and have more lexical
meaning. e ability for preverbs to separate syntactically is a phenomenon known
as tmesis (Booij & van Kemenade, 2003, pp. 1, 88; Diessel, 1999, p. 141; Lehmann,
2015, pp. 104–111; Watkins, 1964). Matthews (2014, p. 318) notes that, “It is per-
haps for this case [of tmesis] that the term [preverb] is most useful.” More timetic
preverbs are typically newer, while more bound preverbs are typically older and
therefore exhibit a greater degree of univerbation. Individual preverbs may also
exhibit divergence, so that lexical and grammatical uses of the same form coexist
synchronically.
In some ways it is useful (though not wholly accurate) to think of preverb + verb
constructions as the syntactic reverse of what are called phrasal verbs or particle
verb constructions in English and other Germanic languages, the primary dier-
ence being the relative order of the verb and the verbal particle/preverb. Phrasal
verbs also show timetic alternations and dierent degrees of compositionality,
© 2018. John Benjamins Publishing Company
All rights reserved
 Daniel W. Hieber
follow some of the same diachronic pathways as preverbs, and contribute lexical
aspect to the verb like preverbs (Los etal., 2012).
Preverbs arise historically from a variety of sources. In Indo-European, pre-
verbs are thought to have originated as independent words, most likely adverbs
(Baldi, 1979). When these proto-preverbs appeared between the direct object and
the verb in transitive OV constructions, it became possible to interpret them as
modifying either the verb or the object, and so they were analyzed as either adverbs
or adpositions respectively (Kuryłowicz, 1964; Watkins, 1964). While this is the
most common source of preverbs crosslinguistically, we will see that Chitimacha
actually did not follow this pathway for its preverbs, and the sources of Chitimacha
preverbs are actually quite diverse. Harris (2003) has documented a similar mul-
tiplicity of sources for preverbs in Udi as well, although the exact mechanisms are
dierent from Chitimacha’s.
Preverbs in Chitimacha are a closed class of nine monosyllabic words that form
a semantic unit with the verb they precede, and convey directional and aspectual
information about the verb. Preverbs constitute the sole exception to the fact that
Chitimacha verbs are morphologically suxing. e list of preverbs, their func-
tions, and their token frequencies (out of 29,028 words in total) is presented in
Table1, along with their most canonical translational equivalent. roughout the
examples in this paper, I gloss preverbs in  , even though they are more
lexical than grammatical.
Each of the preverbs except ni come in plain and reversative pairs, where the
reversative consists of the plain preverb plus a fossilized reversative sux *, e.g.
ʔap ‘here’ and ʔapš ‘back here’. e form of the reversative sux also appears as
-s due to sibilant harmony (e.g. his ‘back to’), though some free variation occurs
between the two forms. Consequently, I do not analyze any word-nal/s/ as the
reversative unless there is additional evidence for the morpheme boundary. Finally,
though both kap and kaːpʼs derive from a root *kaːp ‘up’, their relationship has been
obscured somewhat by historic sound change, to be explained more fully in §§ 3.5
and 3.6 below.
A canonical use of a Chitimacha preverb is shown in (1). In reading the exam-
ples, it will be helpful to keep in mind that (a) verbal person markers only distin-
guish rst (1) and non-rst () person, (b) non-rst person objects are not overtly
marked on the verb, and (c) verbal person marking follows an agent-patient align-
ment system in the rst person and nominative-accusative system in the non-rst
person (Hieber, 2016). First-person axes are agent forms unless otherwise noted.
An appendix of glossing abbreviations is included at the end of this paper.
2
. Transcriptions in the examples follow an Americanist orthography. Notable deviations
from the International Phonetic Alphabet are as follows: < ʼ > = / ˀ /, < c > = / t͡s /, < cʼ > = / t͡sˀ /,
© 2018. John Benjamins Publishing Company
All rights reserved
Chapter2. Category genesis in Chitimacha 
< č > = / t͡ʃ /, < čʼ > = / t͡ʃˀ /, < š > = / ʃ /, and < y > = / j /. Each example is cited along with its source
in Swadesh’s (1939b) text collection, following his system of referencing texts, in which A refers
to speaker Benjamin Paul, followed by the number of the text where the example comes from,
the letter of the paragraph, and the number of the sentence following a period. us A13d.2 re-
fers to the second sentence of the fourth paragraph of the thirteenth text by Benjamin Paul. All
translations in this paper are Benjamin Paul’s (as transcribed by Swadesh) unless given in [square
brackets], in which case they are mine. e interlinear glosses are my own.
Table1. Chitimacha preverbs and their meanings
Preverb Function(s) Translation Token frequency
hi 

‘to’
‘there’
1,298
his 
 


‘back to’
‘back there’
‘again
‘in response’
74
kap 



-
‘beginning’
‘becoming’
‘suddenly’
‘being’
‘up’
775
kaːpʼs ‘back up’ 7
ka  ‘across 1
kas 


‘apar t
‘reverse’
‘back across’
279
ni 



‘doing it’
‘do it!’
‘thing’
‘do w n’
646
ʔap 

‘here’
‘coming’
335
ʔapš 
 



 
‘ab out
‘back here’
‘each other’
‘oneself ’
‘together’
‘coming back here
462
© 2018. John Benjamins Publishing Company
All rights reserved
 Daniel W. Hieber
(1) Panš ʔunkʼu=š kunukʼukap niːk-iʔi.
person one=   be.sick-;
‘ey say a certain person fell sick. (Swadesh, 1939b, A3a.1)
A given combination of preverb + verb may be semantically compositional, like
the examples in (2), or may have shied in meaning and become semantically
non-compositional, like the examples in (3).
(2) a. hi čuw-‘go to’
b. kas čuw- go back, return
c. ni čuw- go down, decrease
d. ʔap čuw- go here, come
e. ʔapš čuw- ‘go about, wander’
(3) a. kas ʔiːkšt- ‘sharpen (tr.)’< ʔiːkšt- ‘turn over’
b. ni wopma- ‘ask (tr./intr.)’
In (2b), the lexeme kas čuw- ‘go back’ can be semantically decomposed into ‘go’ (the
meaning contributed by čuw-) and ‘back’ (the meaning contributed by kas). e
same preverb used with ʔiːkšt- ‘turn over, however, cannot be viewed this way, and
instead the lexeme kas ʔiːkšt- must be analyzed as a holistic, non-compositional unit.
roughout this paper, I will refer to this latter, semantically non-compositional
type of preverb + verb as a lexicalized form, in line with Brinton & Traugott’s (2005,
p. 96) denition of lexicalization as a process where the formal or semantic proper-
ties of a construction are not derivable or predictable from the constituents of that
construction. For example, ni wopma- ‘ask’ in (3b) should be analyzed as a lexical-
ized form because its meaning is no longer recoverable from its component parts
(wop- ‘hear’ + -ma ), and the preverb ni appears regardless of the transitivity
of the clause– the form ni wopma- has become an invariant lexeme meaning ‘ask’. It
is of course sometimes dicult to tell whether a form has lexicalized, but cases like
those in (3) where the meaning is not predictable and the form is largely invariant
are typically easy to discern.
Some but not all preverbs may be timetic, i.e., additional syntactic material may
intervene between the preverb and the verb. is is shown in (4).
(4) Hus waši kiːcti=š we piyi ših =ki hi nam čʼaht-ʼiš-i.
3 hand point=  cane belly = brand hew--;
‘Her thumb (print) is embossed in those cane joints.
(Swadesh, 1939b, A13e.2)
However, these instances of tmesis are limited to just a few specic collocations
and invariable, and therefore are most likely fossilized reexes of a time when
© 2018. John Benjamins Publishing Company
All rights reserved
Chapter2. Category genesis in Chitimacha 
Chitimacha preverbs were syntactically independent from the verb. e phrase hi
nam čʼaht- in (4) historically meant ‘hew a brand into’ (namʔ synchronically means
‘a design or distinctive mark’, while čʼaht- means ‘saw, hew’), but has lexicalized so
that the entire construction now simply means ‘emboss’. Aside from these idiosyn-
cratic cases, Chitimacha preverbs immediately precede the verb with which they
form a lexical unit.
A few preverbs also have additional, non-verbal functions, and so can precede
things like nouns or adjectives, as in (5).
(5) We kaːʼiʔapš keta=nki ni noː-ma-ːš-i.
 three  side=  lay---;
‘ey lay the three [down] side by side. (Swadesh, 1939b, A73b.2)
ough the individual behaviors of Chitimacha preverbs are quite diverse, there are
still a number of reasons for considering them members of a single preverb cate-
gory. In fact, as will be argued below, it is precisely these commonalities that allowed
for schematization across what were originally a diverse group of words. First, the
preverbs share similar phonotactics, all of them monosyllabic with short vowels
except for kaːpʼs. Second, while preverbs are always part of the same intonational
phrase as the verb that follows (Swadesh indicated prosodic phrasing in his texts),
they have not cliticized to the verb. Morphophonological rules do not apply between
the preverb and the verb as might be expected of clitics. In other contexts one sees
/š#hV/  /šV/, but one never sees this with ʔapš + /#h/, for example. Syntactically,
the preverbs only have scope over the main verb, and not the entire verbal phrase
as a clitic might. e third commonality is that all the preverbs have a directional
sense as one of their core meanings, suggesting a semantic basis to the category.
Fourth is that the preverbs participate in the plain/reversative alternation discussed
above (except for ni). Fih, the preverbs always occur in the same syntactic slot and
are mutually exclusive with one another, i.e. in complementary distribution. Only
one preverb can occur with any verb, even when more than one preverb would be
appropriate to the meaning being conveyed. Swadesh (1939d, pp. 147–148; 1946,
pp. 329–330) even describes a set of rules which he calls “preverb displacement”
that determine which of two preverbs will appear when a speaker wants to use a
second preverb with a lexeme that already has one. In this case the preverb that has
become a lexicalized unit with the verb is omitted. For example, the lexicalized verb
his heːčt- ‘meet, join (tr.)’ becomes ʔapš heːčt- ‘meet together’ rather than *ʔapš his
heːčt-. Another shared feature of preverbs is that they frequently form “an essential
part of the verbal lexeme” (Swadesh, 1939d, p. 147), by which Swadesh means that
they form a lexical unit with the verb, and that many of their uses are semantically
non-compositional, as described for Examples(2) and (3) above.
© 2018. John Benjamins Publishing Company
All rights reserved
 Daniel W. Hieber
e nal piece of evidence for a distinct preverb construction is that no other
class of words would be an appropriate alternate home for these nine words, if one
were inclined to separate them into dierent categories. One known source for
preverbs crosslinguistically is preverbal adverbs (Lehmann, 2015, p. 104–105). is
would be surprising for Chitimacha, however, which has no clear class of adverbs.
Almost without exception, adverbial words are minimally bimorphemic, and do not
occur in the same syntactic slot as preverbs. Adverbials occur only clause-initially
or postverbally, and can co-occur with preverbs, oen with a direct object interven-
ing between them. One might also be inclined to treat preverbs as postpositions,
since Chitimacha’s SOV order always places preverbs immediately aer the object
noun phrase. But Example(6) exemplies the way that preverbs may co-occur with
postpositions, even when the two elements exhibit supercially similar meanings.
Since the two words cannot both be postpositions, one of them (the preverb) must
belong to a separate category.
(6) šeːni=nk hup hi ničw-iʔi
pond= to  go.to.water-;
‘he came to the edge of a pond’ (Swadesh, 1939b, A1a.2)
Finally, while preverbs oen imbue aspectual-type meaning to the verb they oc-
cur with, they are not themselves inectional markers of aspect, since they also
co-occur with perfects, perfectives, imperfectives, etc. e aspectual-type semantic
contribution that Chitimacha preverbs make to the verb is therefore best viewed as
a type of lexical aspect (Aktionsart) rather than grammatical aspect. It is common
for preverbs crosslinguistically to contribute this kind of lexical aspect (cf. Los etal.,
2012 for preverbs in Germanic). Example(7) illustrates the aspectual contribution
of the preverb kap, which here functions as an inchoative, while grammatical aspect
is marked by the appearance of -š on the verb.
(7) kap ʔʼiki-ːkʼnaʔa
 rot-- (;)
[‘they have become rotten’] (Swadesh, 1939b, A11c.8)
In sum, if preverbs do not form a category in themselves, it is not clear what other
category they would belong to.
Having given an overview of preverbs generally and in Chitimacha, let us now
examine each in depth, aiming to determine their diachronic trajectories.
© 2018. John Benjamins Publishing Company
All rights reserved
Chapter2. Category genesis in Chitimacha 
. ʔap 
e history of ʔap is the most straightforward of the preverbs. Synchronically ʔap
has both a venitive function meaning ‘coming’ or ‘going here’, as in (8), and an
adverbial demonstrative function meaning ‘here’, as in (9).
(8) We t k hank ʔap nem-naʔa.
then here  cross.water-;
‘en they crossed over to here. (Swadesh, 1939b, A2c.1)
(9) We y t hukʼupanš pinikank ne hank ʔap na.
  Indian just here  (;)
‘at is how the Indian is here. (Swadesh, 1939b, A2c.2)
Notably, Chitimacha has no single unanalyzable verb meaning ‘come’ that might
compete semantically with ʔap. Instead, a construction involving ʔap is used:
(10) We t k kun siksi=s ʔap čuy-i.
then some eagle=  go()-;
‘en an eagle came. (Swadesh, 1939b, A2b.1)
Given that ʔap matches the CVC pattern characteristic of historic verb roots in
Chitimacha, it seems likely that the preverb ʔap has its source in a lexical verb
meaning ‘come’. A diachronic pathway whereby a lexical verb meaning ‘come’ be-
comes a venitive is well-attested crosslinguistically (Heine & Kuteva, 2002, p. 70;
Harris, 2003, pp. 68–69). But what was the mechanism by which the lexical verb
come’ became reanalyzed as a preverb? e most likely candidate is constructions
like that in (11):
(11) ʔaštkankiš ʔhuːta=š ʔap ʔaːy-ʔiš-naʔa.
sometimes 1 boat=  borrow--;
‘Sometimes they come and borrow my boat. (Swadesh, 1939b, A70a.4)
Originally, this would have been a serial verb construction consisting of the un-
inected verb ʔapcome’ followed the fully-inected main verb. Indeed, it is not
uncommon for ʔap to still be translated as ‘come and …’ or ‘come [verb]’. en, via
analogy to and schematization with other preverbs in the making, ʔap would have
been reanalyzed as the venitive preverb. e more strongly adverbial/directional
sense of ʔap meaning ‘here’ or ‘to here’ like in (9) must have therefore been histor-
ically derivative from the venitive sense.
© 2018. John Benjamins Publishing Company
All rights reserved
 Daniel W. Hieber
. ʔapš 
I term the preverb ʔapš a reditive (< Latin reddere ‘to return, send back, give back’),
since its core meaning is ‘coming back’. Diachronically, ʔapš decomposes into ʔap
come’ (see § 3.1) + -š . However, its range of functions is more diverse
than ʔap, and so the diachronic trajectory by which these additional senses devel-
oped must be explained as well.
e meaning of -š as a reversative is apparent from alternations both within the
set of preverbs (compare (12a) and (12b)), and in verbs generally (compare (13a)
and (13b)). Because the appearance of -š outside the preverbal paradigm is limited
to use with verbs, this supports the hypothesis that ʔap was originally a lexical verb
(cf. § 3.1 above).
(12) a. Pešmank=š kunu kap peš-mi-ːkʼtʼut-naʔa.
ducks=    y-- go()-;
‘e ducks have own up and gone. (Swadesh, 1939b, A63a.18)
b. Hesikʼen kaːpʼs nuhčwi-čuy.
again  stand-()\;
‘He will rise up (from his bed) again. (Swadesh, 1939b, A16b.5)
(13) a. Weykš kʼasmi ba-k-te-pa koːš-naʔa.
thus corn at--- call-;
‘ey call it attened corn. (Swadesh, 1939b, A74g.3)
b. kas ba-š-te-
3
 at--
‘fold’ [lit. ‘un-atten back’] (Swadesh 1939c, p.34)
Since we have already seen that ʔap may have both adverbial (‘here’) and venitive
(‘come’) meanings, similar adverbial and venitive meanings for ʔapš can be trans-
parently derived, thus explaining two of the functions of ʔapš. A reditive venitive
use of ʔapš is shown in (14), and a reditive adverbial (‘back here’) use in (15).
(14) ʔam kʼiht-k-š ʔapš ʔehy-i?
what want--  arrive()-;
‘What do you want that you come back?’ (Swadesh, 1939b, A17d.3)
(15) Him te hesikʼen ʔapš ču-ːkʼ-š či-n.
2  again . go()-- (;)-
‘Is that you coming back here again?’ (Swadesh, 1939b, A69c.14)
. is form is taken from Swadesh’s (1939c) dictionary rather than the text collection. It was
elicited from Benjamin Paul as part of a word list.
© 2018. John Benjamins Publishing Company
All rights reserved
Chapter2. Category genesis in Chitimacha 
From these uses, there is a plausible pathway for how the sense of ‘about, randomly’
as illustrated in (16) could have developed. Reditive venitive senses of ʔapš meaning
going and coming (back)’ may have been used ever more guratively until it came
to include meanings like ‘go about’, ‘wander’, and ‘move randomly’.
(16) hiʔniš ʔapš čuː-ma-ːš-či.
just  go()---:;
‘he is simply wandering about’ (Swadesh, 1939b, A7a.7)
e reexive and reciprocal meanings of ʔapš would have then developed from
the ones above. e meanings ‘return’ and ‘come back’ are attested as a source of
reexives in Sanuma (Yanomam, Brazil), resulting from a semantic narrowing from
‘back’ to ‘back to oneself’ (Borgman, 1991, p. 43, as cited in Schladt, 2012, p. 105).
e same process appears to have occurred in Chitimacha, yielding the reexive
use of ʔapš like that in (17).
(17) hus nehe ʔapš kʼet-iʔi.
3 self  kill()-;
‘He killed himself. (Swadesh, 1939b, A3f.7)
Reflexives are themselves a commonly-attested diachronic source for recip-
rocals (Heine & Kuteva, 2002, p. 254), and this is also the case for Chitimacha.
Example(18) shows one such reciprocal use of ʔapš.
(18) We t k kunukʼutep ʔapš ʔaːy-puy-naʔa.
then  re  lend--;
‘en they lend re to each other. (Swadesh, 1939b, A5d.3)
e last sense of ʔapš to be explained is the sociative construction meaning ‘to-
gether’ like that in (19).
(19) Kiš ne ʔapš neːčʼi-mi-ːdi-naːkʼan hesikʼen.
dog even  talk--()-;  again
‘Dogs are not to converse together again. (Swadesh, 1939b, A6b.1)
While it seems plausible that the reditive adverbial ‘back here’ could have developed
into the sociative (since coming back to a place oen includes coming back together
with something le at that place), it is actually reciprocals that are known to be
polysemous with sociatives crosslinguistically (König & Gast, 2008, p. 8). Either
way, a plausible pathway for the sociative use of ʔapš is available.
© 2018. John Benjamins Publishing Company
All rights reserved
 Daniel W. Hieber
. hi 
Like ʔap and most of the other preverbs, hi has directional (20) and adverbial (21)
senses. For hi, however, these exhaust its range of meanings.
(20) we šeːni waʔa=nk hi peš-iʔi.
 pond other=  y-;
‘he ew him toward the opposite side of the pond’ (Swadesh, 1939b, A1c.1)
(21) wetk hi ʔuy-naː
then  arrive()-;-
‘when we got there (Swadesh, 1939b, A65b.9)
e preverb hi does not follow the CVC structure expected for verb roots, and
there are no synchronic reexes in the language to suggest a verbal origin for hi
as there are for some of the other preverbs. ough hardly conclusive, these two
points make a verbal origin for hi less likely. However, the construction in (20) is
suggestive as to its source. Forgács (2004) documents a diachronic pathway for
Hungarian where some preverbs arose from postpositions that were reanalyzed as
belonging with the following verb rather than the preceding noun. Chitimacha hi
shows evidence of following a similar trajectory. In (20) and many other examples
like it, hi is indistinguishable from a postposition, of which Chitimacha has many
(but remember that Example(6) above showed this resemblance to be supercial,
because hi itself can co-occur with other postpositions). Hi contributes a semantic
goal to the meaning of the verb, which thus oen licenses (but does not require) the
presence of an overt object functioning as that goal. A history where hi originated
as a postposition explains this behavior, since hi would have retained some of its
previous constructional properties (namely, the ability to license an argument).
Moreover, just as postpositions typically occur with indirect objects such as goals,
recipients, or beneciaries, Examples(22) and (23) show a case where the presence
of hi imparts a recipient semantics to the NP (22), but in the absence of hi the NP
is interpreted as a patient or perhaps theme (23).
(22) wetk ni ti:kmiš hi koː-naka.
then Governor  call-1
‘we called the Governor. (Swadesh, 1939b, A3.31)
(23) ʔakšuš hečʼin koː-š-naʔa.
cypress holy call--;
‘ey call [i.e. name] them holy cypresses. (Swadesh, 1939b, A9f.2)
It therefore seems that a sequence of [NP PostP] V was reanalyzed as NP [PREV
V], with the result that hi, the preverb in this case, retained some of its earlier prop-
erties, such as the ability to license an overt noun phrase or imply a semanticone.
© 2018. John Benjamins Publishing Company
All rights reserved
Chapter2. Category genesis in Chitimacha 
. his 
e preverb his is another member of a plain– reversative pair, decomposable into
the andative hi (see § 3.3) + reversative -š (realized as /s/ here due to sibilant har-
mony). Some of the senses of his, like ʔapš, can thus be straightforwardly derived
from its non-reversative counterpart. An andative reditive () meaning
is illustrated in (24), and an adverbial reditive ( ) meaning in (25).
(24) ʔunkʼu=š ni tiːkmiš his kow-i
other= Governor  call-;
Another responded to the Governer’ (lit. ‘called back to’)
(Swadesh, 1939b, A4g.7)
(25) hesikʼen his tʼut-k
again . go()-
‘when they went back (there) again (Swadesh, 1939b, A5b.1)
In addition to the adreditive function, his has a repetitive meaning, ‘doing again’,
illustrated in (26). In this example, his appears only with the second, repeated in-
stance of kihci- ‘pound’.
(26) kʼasmi kʼapt-k, […] kihci-ːkʼ, […] hesikʼen his kihci-ːkʼ,
corn take-  pound- again  pound-
‘ey took the corn, […] pounded it, […] pounded it again,
(Swadesh, 1939b, A74e.2)
It is known that morphemes meaning ‘go back (to)’ can develop into iteratives
(Heine & Kuteva, 2002, p. 259), suggesting a pathway from the adreditive to the
repetitive meanings of his. ough not quite a canonical iterative (iterativity in
Chitimacha is typically accomplished through the pluractional marker -ma), this
seems a likely diachronic pathway for the repetitive meaning of his as well.
Swadesh also describes another meaning for his: ‘doing in response’, specically
with verbs of communication (Swadesh, 1939d, p. 152a). It is not clear how this
sense is to be distinguished from reciprocals however. Indeed, the vast majority of
instances of his in the corpus occur with either the verbs of communication ‘answer’
and ‘say’, as in (27), or with the reciprocal-type verbs ‘meet’, and ‘wait for’, as in (28).
(27) we haksikʼam =hiš siksi his nuyt-iʔi
 young_man = eagle  call-;
‘the young man answered the eagle (Swadesh, 1939b, A2b.5)
(28) We t k š siksink his heːčt-iʔi.
then eagle  carry-;
‘en an eagle met him. (Swadesh, 1939b, A1b.1)
© 2018. John Benjamins Publishing Company
All rights reserved
 Daniel W. Hieber
It may be that the semantic distinction between his reciprocals and ʔapš reciprocals
is a matter of aectedness. In the case of ‘meet’ and ‘wait for, both participants
are themes rather than patients. ere is at least one other documented case of a
‘response-reciprocal’ (Camargo, 2007), which supports the potential distinction
between his reciprocals and ʔapš reciprocals, but this merits further investigation.
Of importance here are the diachronic origins of these dierent senses. ese
response-reciprocals can reasonably be assumed to derive from the andative use
of his, likely by a semantic extension of ‘back to’ to mean ‘in response to’, though
I know of no other studies demonstrating a diachronic pathway from reditives or
andatives> responsives.
. kap -
e core directional meaning of kap is ‘up’, glossed here as -, and
exemplied in (29). It also occurs in more gurative uses, as in (30).
(29) pokta=nk kap peš-k
sky=  y-
‘I ew up to the sky’ (lit. ‘ying up to the sky’) (Swadesh, 1939b, A10j.4)
(30) we šuš kap cʼi-tʼi-naʔa-nkʼš
 tree  warm-()-;-
‘they must burn up the tree’ (Swadesh, 1939b, A9e.2)
We can infer the origins of kap through converging evidence from a number of
synchronic stems. e reversative counterpart of kap is kaːpʼs ‘back up’ (where the
nal -s is the reversative; see § 3.6), suggesting that the original form of kap may
have been *kaːp. In support of this hypothesis is the fact that the verb kaːpte- ‘to
sprout, stem’ also appears to derive from a historic root *kaːp meaning ‘up, plus
the intransitive verbal sux -te. e glottalized consonant/pʼ/ in kaːpʼs arose from
the reanalysis of a postvocalic glottal as glottalization on the following consonant,
triggering compensatory vowel lengthening in the vowel preceding. While this
glottal was lost in kaːpte-, and also triggered compensatory lengthening, the historic
glottalization of consonants never occurs before the sux -te, explaining the pres-
ence of a glottalized /pʼ/ in kaːpʼs- and its absence in kaːpte-. However, for reasons
not yet fully understood, compensatory lengthening of vowels was not retained in
all cases. Consider the form kapi ‘seed’: this undoubtedly derives from *kaːp ‘up/to
sprout’ plus the nominalizer -i. So it is not surprising that kap might have lost its
vowel length as well, even if the exact reasons are not fully understood.
It seems then that ‘up’ was one of the earliest meanings of kap. is adverbial
directional function for kap would have set the stage for its reanalysis as a preverb.
© 2018. John Benjamins Publishing Company
All rights reserved
Chapter2. Category genesis in Chitimacha 
But kap underwent numerous other changes as well. A rst development was prob-
ably the extension to punctual changes of state, illustrated in (31) and (32), and
seen most frequently with verbs like ‘die’ and ‘stop’. ese uses would have been an
extension of the gurative use of ‘up’ seen above.
(31) We t k we ʔašinčʼatʼa=š kap nuːp-iʔi.
then  old_man=  die()-;
‘en the old man died. (Swadesh, 1939b, A16c.4)
(32) Tutk kunukʼuhunks ni kʼuštiš kap čip-iʔi.
then  3 food  nish-;
‘eir food ran out. (Swadesh, 1939b, A3b.2)
Detges (2004) describes how movement, including verbs meaning ‘jump’ or ‘leap’,
can be a robust source for inchoatives with examples from Indo-European: “If an
agent moves to some place with the intention of carrying out some action there,
then she is visibly making a gesture which will take her to the beginning of this
action” (Detges, 2004, pp. 213.). e act of getting up is likewise an indication that
an action is about to take place, and so we see that the ‘up’ meaning of kap developed
into inchoative and inceptive senses, in line with Detges’ cognitive perspective:
(33) kaːkwa-ki ʔašt ʔuči ːkʼšpanš ne kap nacpik-mi-naʔa
know 1;  how do- person even  begin--;
‘I do not know how people started up (Swadesh, 1939b, A1d.4)
(34) kap paːkine-ki-čuː
 be_tired-1;-()-
‘if I get tired’ (Swadesh, 1939b, A2b.8)
Note that all instances of the inceptive use of kap in the corpus (i.e. those translated
with ‘begin’) co-occur with the lexical verb nacpik- ‘begin’, so it is not clear that this
should even be considered a discrete function for kap. Inchoative uses, however,
occur with a variety of verbs.
In opposition to its punctual, change of state use, kap may also be used with
nouns and deverbal adjectives to impart a durative stative reading, as in (35). is
was probably an extension of the inchoative meaning, so that ‘became happy’ took
on a perfect reading, where the change of state is viewed as still relevant to the pres-
ent. In fact, these stative readings are most common in perfect aspect constructions,
as in (35). In this passage, an event occurred in the prior clause which made the
people happy, and so (35) conveys a change of state (becoming happy) that then
endured for some time.
© 2018. John Benjamins Publishing Company
All rights reserved
 Daniel W. Hieber
(35) We t k we panšk kap šeški-ːkʼna-ʔuy-naqa.
then  people  be_happy-- ()--;
‘e people were happy. (Swadesh, 1939b, A11c.14)
Without kap, the change of state meaning in this construction is lost, leaving just
a durative meaning.
. kaːpʼs 
e preverb kaːpʼs occurs only 7 times in the corpus, each time translated as ‘rise
up’ or ‘get up’ in a situation where the sense of ‘back up’ is strong:
(36) kaːpʼs kay-mi-ːtʼi-nan.
 rise--()-;
‘they will rise up [from the dead]. (Swadesh, 1939b, A11c.10)
kaːpʼs derives from *kaːpʼ + -s , and is related historically to its
non-reversative counterpart kap through their common root *kaːp ‘up, as was de-
tailed in § 3.5. kaːpʼs has no other documented functions besides the superreditive
‘returning up/back up’.
. ka 
e preverb ka is a hapax legomenon appearing only once in the corpus, and was
not noted by Swadesh in any of his writings on the language. However, its meaning
in (37) is exactly what would be expected if ka were the non-reversative counterpart
of kas (see § 3.8 for the meanings and functions of kas).
(37) we kimuš ney=up ka nenšt-k
 branch land=to  take.across.water-
‘[they] brought the limb to land’ (lit. ‘taking across water’)
(Swadesh, 1939b, A9c.8)
It is clear from the broader discourse context of this example that the limb is being
brought across the water, the expected meaning for ka, since the characters are
crossing a lake, and because the verb root nen- literally means ‘on/over water’.
ere is also independent evidence for a historic root meaning ‘across’, which
not only conrms the analysis of ka, but explains its diachronic trajectory: the
verb kaːkte- ‘extend across’ decomposes into the historic root *kaːk + -te , and
*-kaːk very likely derives from *kaʔ + * k , where both the stative *-k and
the pattern of loss of a glottal accompanied by compensatory lengthening before
© 2018. John Benjamins Publishing Company
All rights reserved
Chapter2. Category genesis in Chitimacha 
the addition of a historic stem sux are well-attested (see also the change from
kaːpkap above). In short, *kaʔ was very likely a historic root meaning ‘across’ or
go across, which served as the diachronic source of the preverb ka. is pathway
also nicely parallels that for ni (cf. § 3.9 below), where a historic /CVʔ/ root with a
directional meaning (*niʔ ‘(go) down’) developed a preverbal function.
. kas 
e core directional meaning of kas is a translative reditive or transreditive, mean-
ing roughly ‘(going) back across’. Of the reditive preverbs, kas seems to be the un-
marked form for expressing ‘back’ or ‘returning’, and very frequently gets translated
with these two English verbs, more so than the other reditives:
(38) ʔašt kas tʼuːt-ʼš-naʔa
how  go()--;
‘How are you going back?’ (Swadesh, 1939b, A3f.1)
kas derives from the preverb ka  + -s . e dislative (‘(go-
ing) apart’) sense of kas is analogous to the adverbial uses of other preverbs, where
the semantics of the preverb shis from the type of movement to the result of the
movement. In this case, the resulting position is ‘apart’. is sense is illustrated
in(39).
(39) we tep kas heːčt-k
 re  carry-
‘[they] raked the re apart’ (lit. ‘carrying the re apart’)
(Swadesh, 1939b, A74p. 4)
ere was also a semantic extension whereby the act of going back to a place was
construed more generally as an act of reversing a process, and so kas also came to
have a general reversive meaning. Example(40) is an instance of this.
(40) we panš pinikank ʔašinčʼatʼa=š panš kas tey-i-nki
 Indian old.man= person  become()-;-
‘aer the old Indian turned back into a person (Swadesh, 1939b, A28d.3)
. ni 
Of the preverbs, the functions of ni are the most dicult to reconcile with one
another. Ni occurs with non-verbal elements more frequently than any other pre-
verb by far, appearing in numerous deverbal nominalizations (41) and noun-noun
compounds (42).
© 2018. John Benjamins Publishing Company
All rights reserved
 Daniel W. Hieber
(41) ni tʼap-kʼi
 be.dark-
dark/darkness (Swadesh, 1939b, A5f.3)
(42) a. šaːpniš ni poʔ
rattlesnake  plant
‘rattlesnake medicine (Swadesh, 1939b, A75g.4)
b. hi čʼiːpampa ni čʼah
pet  bird
‘pet bird (Swadesh, 1939b, A12a.6)
c. ni šaʔ
 mouth
‘voice/language’ (Swadesh, 1939b, A50a.7)
e most canonical function of ni when used with verbs is as a detransitivizer. is
function is evidenced by the many pairs of examples like those in (43).
(43) a. ʔʔnuːp kʼas-ka-nki-š
1 1 potato plant---
‘when I planted my potatoes’ (Swadesh, 1939b, A59b.1)
b. hesikʼen ni kʼas-mi-naʔa
again  plant--;
‘again they planted’ (Swadesh, 1939b, A3b.6)
ere are however very many exceptions to this pattern. Transitivity in Chitimacha,
like all languages (Hopper & ompson, 1980), is not at all binary, and instead de-
pends on a variety of factors such as the lexical aspect of preverbs, valency eects
from preverbs, the presence of certain transitivizing or detransitivizing verb suxes,
grammatical aspect on the verb, the presence of an overt noun phrase, nominal case
marking, the presence of the verbal pluractional sux, and the choice of agent
versus patient prexes. Each of these features is in turn motivated by discourse
(Hopper & ompson, 1980) and the event construal of speakers (see especially
Martin, 2000 for a discussion of event perspective as it relates to valency). As such,
even when the preverb ni is present, the clause may be highly transitive in other
ways. is can be seen in (44b), where the presence of an overt direct object noun
phrase does not prohibit the appearance of ni ((44a) is included for comparison).
(44) a. we čʼaːšaʔa=š tuč-iʔi
 rice= cook-;
‘he cooked the rice’ (Swadesh, 1939b, A15e.5)
b. ʔušk čʼaːšaʔani tuč-mi-naka-š
1 rice  cook--1-
‘when we cooked rice’ (Swadesh, 1939b, A74a.6)
© 2018. John Benjamins Publishing Company
All rights reserved
Chapter2. Category genesis in Chitimacha 
Because of this seeming inconsistency, Swadesh himself states that “the force of the
usage [of ni] is completely unclear” (Swadesh 1939d, p. 154). John R. Swanton, how-
ever, working with the same speaker two decades prior to Swadesh, suggested that
ni actually meant ‘something’, and treats it as the direct object of the verbs it occurs
with (Swanton, 1920, p. 10). While analyzing ni as a direct object is problematic be-
cause ni can co-occur with a full direct object noun phrase (as in (43b)), an analysis
of ni as a synchronic reex of a historic pronoun meaning ‘thing’ provides a neat
account of the data. It explains why ni occurs in nominalizations like (41) and (42),
while also having a detransitivizing eect in cases like (43). In the detransitivizing
cases, ni probably originally served as the direct object, and later was reanalyzed
as a detransitivizing preverb. Because the direct object slot was already being lled
by ni at the point in time when ni was reanalyzed as a preverb, transitive construc-
tions like ni kʼas-, which would have originally been parsed as ‘to plant something’
(transitive verb + direct object), were reanalyzed as a single lexical unit meaning
‘to plant’ (intransitive, with the semantic implication, provided by the preverb, that
there is some specic thing being planted).
4 is process whereby preverbal nom-
inals are incorporated into the verb and in doing so aect the verb’s transitivity is
most well documented in cases of historical noun incorporation (Mithun, 1984).
A nal extension from the ‘thing’-related senses of ni is to an imperative marker.
ni is oen redundant in these cases, since the verbs it occurs with are also marked by
the imperative sux -(ʔ)a or a special imperative stem, as in (45) and (46) respec-
tively. In addition, imperative ni oen co-occurs with another imperative particle
huš, also seen in (46), whose meaning is unknown. However, in cases like (47), ni
is the only formal marking of the imperative.
(45) kahpi ni kaːčt-ʼa
coee  drink-
drink some coee!’ (Swadesh, 1939b, A29c.1)
(46) Pušinkank huš ni pe.
quiet   ()
‘Remain still!’ (Swadesh, 1939b, A17g.10)
. One reviewer wonders whether the nominalizing and detransitivizing functions of ni could
be considered part of the same development, since deverbal nominalizations usually involve a
reduction in valency. is does not seem to be the case for Chitimacha. Verbs used with ni do
not exhibit any other nominal-like behaviors. ey may be used as main or subordinate clauses,
and may be nite or non-nite, independent of the presence of ni. Plausibly, frequent enough use
of ni with nouns could, via analogy, cause the reanalysis of ni + verb constructions as nominals,
but I have yet to nd any evidence to this eect.
© 2018. John Benjamins Publishing Company
All rights reserved
 Daniel W. Hieber
(47) Ni way-ma sa šaːhken.
 weave-  basket
‘Weave that basket!’ (Swadesh, 1939b, A13a.4)
is imperative function could have plausibly developed from a reanalysis of
phrases like (47), which would have originally meant ‘weave it, that basket’, to a
more general imperative marked by ni. Eventually ni could have extended its dis-
tribution beyond transitive verbs to contexts like (46).
e only other major sense of ni is, to the best of my knowledge, unrelated to
the senses above. is is the sense of ‘down’, as seen in (48)–(50).
(48) kuːni čuw-a=nki
water  go-;=
‘When the water went down (Swadesh, 1939b, A10e.1)
(49) Te w e we kuːkni čuy-i.
but  water  go-;
At any rate the water went down. (Swadesh, 1939b, A62b.5)
(50) We t k we šuš ni tey-p-iʔi.
then  tree  sit()--;
‘He put the tree down. (Swadesh, 1939b, A12b.3)
ere is good synchronic evidence that the ‘down’ meaning of ni is quite old, so
that it is unlikely that the ‘down’ meaning developed from the ‘thing’ meaning. Two
forms in particular suggest that ‘down’ as a meaning of ni is signicantly old, since
it is buried behind another fossilized sux, *-h, within the verb root. e historical
internal morphology of these two forms is shown below.
(51) niːhkup
*niʔ-*-h -k hup
go.down in  to
‘dow n’
(52) nehčwa-
*ne/ni *-h -čwa
down in move.vertically
‘walk down (also numerous other verbs with an element *neh-)
ese historic morphemes can be reconstructed from other forms as well. Since I am
not aware of any documented pathway whereby an element meaning ‘thing’ came to
mean ‘down’, it seems the best explanation for the use of ni as both a nominalizer/
detransitivizer and a directional meaning ‘down’ is a diachronic merger between
what were originally two near-homophonous forms: *ni ‘thing’ and *ni(ʔ)‘down’.
© 2018. John Benjamins Publishing Company
All rights reserved
Chapter2. Category genesis in Chitimacha 
. e constructionalization of Chitimacha preverbs
Having examined the probable diachronic pathways by which the various senses
of Chitimacha preverbs may have developed, we are now in a position to deter-
mine how the preverb category could have emerged from a set of forms with such
disparate histories. e key, I argue here, is that each proto-preverb independently
underwent a series of micro-changes (constructional changes) that happened to
converge on a shared set of properties. Speakers, recognizing these shared proper-
ties, abstracted away from the various constructions to recognize the existence of a
new schema. en began a process of reanalysis whereby the morphosyntactic prop-
erties of each of the proto-preverbs was brought in line with their newly perceived
function. I term this process schematization, a mechanism whereby constructions
undergo reanalysis as a result of the recognition of a new cross-constructional
schema by speakers.
e development of Chitimacha preverbs is, conceptually speaking, similar to
a much-discussed case in the grammaticalization literature: the development of
English auxiliaries (Heine, 1993; Hopper & Traugott, 2003, pp. 55–58; De Smet,
2009, p. 1751; Roberts & Roussou, 2003, pp. 36–48). I take both to be cases of
category genesis. Indeed, De Smet (2009, p. 1751) poses the same question that I
have posed here:
English at some point introduced auxiliaries, so conceivably there must have been
a rst auxiliary, but how could the rst English auxiliary be analysed as an auxiliary
without drastic reanalysis, given that analogically-based categorial incursion is
impossible in the absence of other auxiliaries? So how could auxiliaries ever emerge
without a rst auxiliary? (De Smet, 2009, p. 1751, emphasis added)
De Smet’s answer is to appeal to analogy, in line with his broader endeavor to show
that reanalysis may be reduced to more fundamental mechanisms, analogy being
foremost among them. In answer to his question, he states,
e answer, I believe, is that, paradoxically, the rst English auxiliary could not
be analysed as an auxiliary until there was a second one. Before that time, the
‘auxiliary’ would have been an under-analysed and grammatically isolated chunk
of language that had undergone both gradual category-internal change and auto-
mation. Only when another such chunk developed, language users could perceive
a similarity between the two. At that point a category ‘auxiliary’ arises, which,
however, entails no more than a perceived similarity. (De Smet, 2009, p. 1751)
Certain aspects of De Smet’s analysis are intuitively appealing, in particular the
idea that a perceived similarity between forms may arise that serves as the motiva-
tion for the new category. However, I nd De Smet’s ‘paradoxical’ explanation too
© 2018. John Benjamins Publishing Company
All rights reserved
 Daniel W. Hieber
conceptually circular to be satisfying. e main problem is that his treatment of
analogy is unidirectional: it only allows for form A to become more like form B, or
form B more like form A. What is needed is a process that makes both form A and
form B into a third form C, without form C existing prior to the change.
A more robust denition of analogy is adopted by Traugott & Trousdale (2013).
First, they distinguish between analogical thinking, which is the recognition of pat-
terns of similarities between meanings and forms, and analogization, “a mechanism
or process of change bringing about matches of meaning and form that did not
exist before” (Traugott & Trousdale, 2013, p. 38; see also Traugott & Trousdale,
2010, p. 38). Analogical thinking and pattern matching may or may not lead to
analogization. Dened this way, analogization simply brings forms in line with
their meanings, where those meanings have changed as the outcome of analogical
thinking. Analogization is always realized through reanalysis, and in cases of anal-
ogization that reanalysis is motivated by analogical thinking (Traugott & Trousdale,
2010, p. 38); other kinds reanalysis are driven by other motivations. Under this
understanding, form A and form B may both undergo reanalysis to better align
with their newly perceived meaning, creating a new form C.
In the case of Chitimacha preverbs and category genesis generally, that per-
ceived similarity is necessarily abstract and therefore schematic, since it holds
across disparate constructions. While all schemas necessarily cut across dierent
constructions, schemas that arise in the process of category genesis are especially
cross-constructional because they link together constructions that were not previ-
ously recognized to have much, if anything, in common (as was the case with the
proto-preverbs in Chitimacha). More typical cases of analogization involve the
extension of a preexisting schema to forms that are already quite similar and share
many properties in common. Put another way, the schemas involved in classic
analogization are abstractions over constructions that are already part of a tightly
connected constructional network. e schemas that arise in the process of cate-
gory genesis, however, link together nodes in the constructional network that were
previously only weakly connected. us prototypical analogization is more about
the extension or change of preexisting schemas to encompass additional forms,
while prototypical category genesis involves the creation of new schemas that hold
between previously unconnected forms.
It should be immediately noted that this distinction is impossible to uphold in
principle, since all schematic changes could be considered new schemas, and new
schemas are built by abstracting away from the properties of existing ones. But
there does seem to be a useful sense in which the schemas involved in analogical
extension and category genesis dier, if only in degree rather than kind. I agree with
Traugott & Trousdale (2013, p. 58) that “no construction is entirely new.” As one
© 2018. John Benjamins Publishing Company
All rights reserved
Chapter2. Category genesis in Chitimacha 
reviewer rightly notes (and I paraphrase slightly), “schemas are abstractions over
sets of constructions, and can not pre-exist constructions. us I do not claim that
the process of category genesis creates schemas de novo. Quite the opposite: while
category genesis appears prima facie to suggest the possibility of a schema arising
without members, in actuality the members that constitute it are the disparate
constructions across which a pattern has been recognized. is is why I emphasize
the cross-constructional nature of the schemas that arise in category genesis: they
cross-cut other schemas, and link previously disconnected constructions. At the
same time, I do not think that reanalysis never results in a totally new structure,
contra Fischer (2007, pp. 123–124) and Itkonen (2005, pp. 110–113). Category gen-
esis seems the perfect counterexample to this claim. While the schema that arises
during category genesis is grounded in the properties of existing schemas, and
thus not entirely new, the structural changes that the new schema instigates may
be completely new, as is the case with Chitimacha preverbs.
With these caveats in mind, I suggest that category genesis is best viewed as a
process of reanalysis motivated by the recognition of a new, cross-constructional
schema. It is a process similar and parallel to analogization, which is a kind of rea-
nalysis motivated by the recognition of a new analogy between existing construc-
tions. In the same way that Traugott & Trousdale (2010, p. 38) distinguish between
analogical thinking (the motivation) and analogization (the mechanism), it seems
useful to introduce the dierence between schema recognition (the motivation),
whereby speakers attend to patterns that hold across constructions, and schemati-
zation (the mechanism), the process whereby the various forms that participate in
the pattern come to align morphosyntactically with the newly recognized schema.
Like analogical thinking, schema recognition enables but does not entail schemati-
zation; and like analogization, schematization is always realized through reanalysis.
Let us now turn to the specics of category genesis as exhibited by Chitimacha.
Since constructions are pairings of form and function, and constructionalization is
the creation of new form-meaning pairings (Traugott & Trousdale, 2013, p. 22), it
is useful to characterize the development of Chitimacha preverbs with a construc-
tional schema. e novel form that Chitimacha preverbs had was the preverbal
syntactic position, and their novel meaning was their contribution of lexical aspect
and directionality to the semantics of the verb. is is schematized in (53).
(53) [[ij] [lexical aspect/directionalityi– j]]
Each preverb, however, has its origins in a construction very dierent from this
one. A simplistic representation of the original constructions for each preverb is
given in Table2.
© 2018. John Benjamins Publishing Company
All rights reserved
 Daniel W. Hieber
Table2. e proto-preverb constructions
Preverb Original construction
ni [[  ] [‘thing’  ]]
ʔap [[][‘come’ ]]
ʔapš [[][‘come back]]
hi [[[ ] ] [[ ‘to’] ]]
his [[[ ] ] [[ ‘back to’] ]]
kap [[ []] [‘up]]
kaːpʼs[[ []] [‘back up’ ]]
ka [[ []] [‘across’ ]]
kas [[ []] [‘back across’ ]]
Each preverb then underwent a series of micro-changes (constructional changes)
in both meaning and form, which were outlined in § 3 above. Figure1 is a seman-
tic map summarizing the semantic changes in the preverbs over time, with older
meanings positioned towards the le, and newer ones towards the right.
Of particular note is the point in time when each of the proto-preverbs had
developed a directional sense as one of its meanings, represented by the large rec-
tangular box in Figure1. is allowed for a process of schematization across the
dierent forms. e light paradigmaticity that existed at this point was the basis for
the recognition of the new schema, setting the stage for the subsequent schematiza-
tion via reanalysis that was to follow. e preverb ni is given two pathways in this
schematic representation, indicating the apparent independence with which the
two senses developed. e ‘thing’ sense of ni is excluded from the box because this
sense would not have contributed to the semantic schema. e ‘thing’ sense of ni
does however contribute to the syntactic component of the preverb schema, since
this sense of the proto-preverb was preverbal like all the others.
e changes in meaning to the le of the box in Figure1 brought about enough
similarities across the proto-preverbs that speakers began to recognize a new seman-
tic schema. ese ‘preparatory’-type changes are termed pre-constructionalization
constructional changes by Traugott and Trousdale and are said to “enable or ‘feed’
constructionalization” (Traugott & Trousdale, 2013, p. 27). is is exactly what we
have seen here: the pre-constructionalization constructional changes– the semantic
shis in the meanings of the proto-preverbs– are what enabled the schematiza-
tion of preverbs to take place. Traugott and Trousdale hypothesize that types of
pre-constructionalizational changes might include expansion of pragmatics, se-
manticization of that pragmatics, mismatch between form and meaning, and some
small distributional changes. e present study supports this view. For example,
we have seen an expansion and then semanticization of pragmatics in the way that
© 2018. John Benjamins Publishing Company
All rights reserved
Chapter2. Category genesis in Chitimacha 
the meaning of ʔapš expanded from a reditive venitive to more gurative senses
like ‘wander’ and ‘move randomly’. Aer rst acquiring an implicature where ‘(go
and) come back’ pragmatically implied ‘wander’ or ‘move randomly’, that prag-
matic interpretation then became conventionalized until it was simply part of the
meaning of the word itself.
An open question is to what extent the aspectual contribution of the preverbs
was also part of this schema, or whether their aspectual functions developed later,
only aer they came to be associated more strongly with the verb. If the proto-
preverbs did develop such aspectual senses before their constructionalization into
preverbs, this would have been one more commonality across the dierent forms
that would have contributed to speakers’ analogization. e fact that the proto-
preverbs constructionalized as preverbs rather than as preverbal adverbs suggests
‘thing’
‘down/under’


‘co m e’ ‘here’‘to here
‘come back’ ‘back to here’ ‘back here’ ‘randomly’ 
 ‘to there ‘th ere’
‘downward’ ‘dow n’

‘back to’ ‘r es ponse’
‘up’ ‘go up’ ‘up’  


‘go back up’ ‘back up’
‘go across’ ‘across
‘go back across’ ‘back across’
ʔap
ʔapš
hi
his
kap
kaps
ka
kas
ni
Figure1. Semantic shis in the history of Chitimacha preverbs
© 2018. John Benjamins Publishing Company
All rights reserved
 Daniel W. Hieber
that this was in fact the case. If it was primarily the directional meanings of the
proto-preverbs that schematized and not their aspectual meanings, there may have
been less of a semantic connection to the verb, and less reason to reanalyze the
proto-preverbs as being syntactically bound. If, as I suspect, the aspectual mean-
ings were already part of the schema, however, the semantic connection to the
verb would have been stronger, inclining speakers towards reanalyzing the proto-
preverbs as syntactically-bound preverbs.
e proto-preverbs had also become schematic in another way, namely that a
high frequency of their occurrences were appearing in a position immediately prior
to the verb. is fact is apparent from Table2. Looking at the le-hand side of the
constructional schema for each preverb (the form portion of the form-meaning
pairing), the one commonality across all the constructions is that the proto-preverb
immediately precedes the main verb. In their syntactic properties, too, we have seen
in the sections above that certain pre-constructionalization constructional changes
took place, such as the change from ʔap as an independent verb to one that is more
serial in nature and syntactically bound to the main verb, or the reanalysis of hi
from modifying the object to modifying the verb.
Given the now high degree of schematicity across each of the proto-preverb
constructions, it then became possible for speakers to reanalyze the proto-preverbs
to bring their syntactic form in line with their schematic meaning. Since each of the
proto-preverbs was viewed as belonging to a schema in which the proto-preverbs
had a directional or aspectual semantics tightly tied to the verb, their syntactic
status came to represent this fact, and they became tightly tied to the verb as well.
Over time, the preverbs by and large lost the ability to undergo tmesis and separate
syntactically from the verb.
Another change thought by Traugott and Trousdale to accompany construction-
alization is a shi in the degree of compositionality of the construction, and we have
seen this process at work here as well. While each of the preverbs began as composi-
tional, it was noted in § 3 that many uses of preverbs are lexicalized and semantically
non-compositional. Interestingly, this varies drastically from preverb to preverb.
Hieber (2014), using the number of headwords in Swadesh’s (1939c) dictionary that
contain each preverb as a rough heuristic of the compositionality of that preverb,
shows that the preverbs vary strongly in their compositionality, independent of
their frequency. While hi occurs 1298 times in the corpus and kap 775 times (the
top 2 most frequent preverbs), hi participates in just 30 lexicalized preverb + verb
combinations, whereas kap participates in 183. While it would be inaccurate to say
that the preverb + verb construction as a whole is always non-compositional, some
of the individual preverb + verb constructions certainlyare.
© 2018. John Benjamins Publishing Company
All rights reserved
Chapter2. Category genesis in Chitimacha 
Finally, any additional senses in Figure1 that developed aer the schematiza-
tion/constructionalization of the preverbs should be considered post-construction-
alization constructional changes (Traugott & Trousdale, 2013). Likely candidates for
these types of changes are the development of the reexive and reciprocal senses
of ʔapš.
. Conclusion
is paper has addressed the question ‘How could Chitimacha preverbs have de-
veloped as a new category in the language, if there were no pre-existing preverbs
on which to analogize?’. e diverse origins of the preverbs makes it challenging to
analyze the creation of the preverb category as a unied process. It seems that one
would need to provide a dierent diachronic pathway for each preverb, but this
still leaves the diculty of explaining how it is that these dierent pathways just
happened to converge on the same set of properties.
e answer proposed in this paper is that what enabled the creation of the
preverb category was the recognition of a generalization across a variety of forms
that all happened to share similar properties, specically, a preverbal syntactic po-
sition and a directional semantics. It is not that speakers had nothing on which
to abstract over, but rather that the abstraction was across all the proto-preverbs.
Speakers recognized a schema that included light paradigmaticity (pairs of plain
and reversative forms), directional semantics, and preverbal syntax. e preverbs
then underwent reanalysis and changes in form to better match the schema they
were seen to be members of, thereby converging in many of their formal properties
as a result, a process I term schematization. is is a slightly dierent process than
analogization: each preverb was not changing to become more like any particu-
lar other preverb; rather, all the preverbs were changing to become more like the
schema to which they all belonged.
One question that remains, and one that the synchronic nature of the
Chitimacha corpus unfortunately does not allow us to answer with certainty, is to
what extent the preverbs arose in tandem versus at dierent points in time. Could
it be that rst there were smaller subschemas that subsequently attracted other
constructions until they developed into preverbs? Or did the proto-preverbs each
develop their preverb-like properties independently, and then form the new preverb
category all together? Most likely neither is a fully accurate characterization. To the
extent that the preverbs developed via a succession of analogizations across already
similar micro-constructions, we can simply call the development of preverbs a
© 2018. John Benjamins Publishing Company
All rights reserved
 Daniel W. Hieber
gradual convergence of mutually-reinforcing changes. But given the rather dispa-
rate origins of certain preverbs, it seems unlikely that speakers would have formed
an analogy between them without there having rst been at least some similarities
in place. A postposition meaning ‘to’ (the origin of hi) is, aer all, not in many ways
similar to a serial verb meaning ‘come’ (the origin of ʔap). e more similarities
each of the preverbs developed independently of each other, the more motivation
there would have been for the rst analogies between them. To the extent that the
rst micro-constructions to participate in the preverb schema were dissimilar to
each other, this argues for the recognition of schematization as a mechanism of
constructionalization alongside analogization.
Regardless of the answer to the above questions, and the status of schematiza-
tion as a distinct process, the constructional approach to category genesis provides
a unied account of the development of Chitimacha preverbs. While there is no
one pathway that holds for all of the preverbs, there is a single constructionaliza-
tion process at work that appropriately applies to all of them. Moreover, it has been
demonstrated here that, at least for Chitimacha preverbs, category genesis can be
productively treated as just a special case of category change– no additional the-
oretical machinery is required to explain category genesis above and beyond that
posited to explain category shis more generally.
is paper also adds to the growing body of literature on reconstruction from
a constructional perspective (see especially Barðdal etal., 2015), in particular
those that focus on syntactic reconstruction (Barðdal & Eythórsson, 2012a, 2012b;
Barðdal & Smitherman, 2013; Barðdal etal., 2013; earlier work includes Harris &
Campbell, 1995; Gildea, 1992, 2000), which necessarily looks at entire construc-
tions rather than individual morphemes and lexemes alone. In this paper it was
also necessary to examine the syntactic context of the proto-preverb constructions,
in order to understand what formal features were similar across all of them. It was
shown that a preverbal syntax was a crucial property contributing to the schema
that developed over the proto-preverbs. Moreover, because the available evidence
is entirely language-internal, I hope to have shown that constructional approaches
to diachrony can be fruitfully applied to the task of internal reconstruction as well.
In conclusion, diachronic construction grammar is suciently robust to handle
what seems prima facie like an extreme and dicult case– the genesis of entirely
new categories out of a collection of otherwise unrelated forms within a language.
© 2018. John Benjamins Publishing Company
All rights reserved
Chapter2. Category genesis in Chitimacha 
Acknowledgements
An earlier version of this paper was presented at the Silver Anniversary Conference at the Uni-
versity of California, Santa Barbara, October 11, 2014. Many thanks to the audience members
there for their feedback. I am also grateful to Marianne Mithun and Jack Martin for their feed-
back and guidance, Tim ornes and one other anonymous reviewer for their helpful com-
ments, and Muriel Norde, Kristel Van Goethem, Evie Coussé, and Gudrun Vanderbauwhede for
their careful editorial work and review. is chapter is much improved for all of these contribu-
tions. is chapter also would have never seen the light of day were it not for nancial support
from the Graduate Division of the University of California, Santa Barbara, the Department of
Linguistics, and a Graduate Research Fellowship from the National Science Foundation (Grant
#1144085). All mistakes and shortcomings are of course wholly my own.
References
Baldi, P. (1979). Typology and the Indo-European prepositions. Indogermanische Forschungen
[Indo-European Researches], 84, 49–61.
Barðdal, J., Bjarnadóttir, V., Danesi, S., Dewey, T. K., Eythórsson, T., Fedriani, C., & Smitherman,
T. (2013). e story of “woe.Journal of Indo-European Studies, 41(3–4), 321–377.
Barðdal, J., & Eythórsson, T. (2012a). Reconstructing syntax: Construction grammar and the
comparative method. In H. C. Boas, & I. A. Sag (Eds.), Sign-based construction grammar
(pp. 257–308). Stanford: CSLI Publications.
Barðdal, J., & Eythórsson, T. (2012b). Hungering and lusting for women and eshly delicacies:
Reconstructing grammatical relations for Proto-Germanic. Transactions of the Philological
Society, 110(3), 363–393. doi: 10.1111/j.1467-968X.2012.01318.x
Barðdal, J., Smirnova, E., Sommerer, L., & Gildea, S. (Eds.). (2015). Diachronic construction gram-
mar. Amsterdam/Philadelphia: John Benjamins. doi: 10.1075/cal.18
Barðdal, J., & Smitherman, T. (2013). e quest for cognates: A reconstruction of oblique subject
constructions in Proto-Indo-European. Language Dynamics & Change, 3(1), 28–67.
doi: 10.1163/22105832-13030101
Bergs, A., & Diewald, G. (2008). Introduction: Constructions and language change. In A. Bergs,
& G. Diewald (Eds.), Constructions and language change (pp. 1–22). Berlin/New York: De
Gruyter Mouton. doi: 10.1515/9783110211757.1
Booij, G., & van Kemenade, A. (2003). Preverbs: An introduction. In G. Booij, & J. van Marle
(Eds.), Yearbook of morphology 2003 (pp. 1–120). New York: Kluwer Academic Publisher
s. doi: 10.1007/978-1-4020-1513-7_1
Borgman, D. (1991). Sanuma. In D. C. Derbyshire, & G. K. Pullum (Eds.), Handbook of
Amazonian languages (Vol. 2, pp. 15–258). Berlin/New York: De Gruyter Mouton.
Brinton, L. J., & Traugott, E. C. (2005). Lexicalization and language change. Cambridge: Cambridge
University Press. doi: 10.1017/CBO9780511615962
Camargo, E. (2007). Reciprocal, response-reciprocal and distributive constructions in Cashinahua.
In V. P. Nedjalkov (Ed.), Reciprocal constructions (pp. 1865–1914). Amsterdam/Philadelphia:
John Benjamins. doi: 10.1075/tsl.71
© 2018. John Benjamins Publishing Company
All rights reserved
 Daniel W. Hieber
De Smet, H. (2009). Analysing reanalysis. Lingua, 119(11), 1728–1755.
doi: 10.1016/j.lingua.2009.03.001
Detges, U. (2004). How cognitive is grammaticalization? e history of the Catalan perfet peri-
fràstic. In O. Fischer, M. Norde, & H. Perridon (Eds.), Up and down the cline: e nature of
grammaticalization (pp. 211–228). Amsterdam/Philadelphia: John Benjamins.
doi: 10.1075/tsl.59.11det
Diessel, H. (1999). Demonstratives: Form, function, and grammaticalization. Amsterdam/
Philadelphia: John Benjamins. doi: 10.1075/tsl.42
Fischer, O. (2007). Morphosyntactic change: Functional and formal perspectives. Oxford: Oxford
University Press.
Forgács, T. (2004). Grammaticalisation and preverbs. Acta Linguistica Hungarica, 51(1–2), 45–84.
doi: 10.1556/ALing.51.2004.1-2.3
Gildea, S. (1992). Comparative Cariban morphosyntax: On the genesis of ergativity in independent
clauses (Doctoral dissertation). University of Michigan, Ann Arbor.
Gildea, S. (Ed.). (2000). Reconstructing grammar: Comparative linguistics and grammaticalization.
Amsterdam/Philadelphia: John Benjamins. doi: 10.1075/tsl.43
Givón, T. (2000). Internal reconstruction. In S. Gildea (Ed.), Reconstructing grammar:
Comparative linguistics and grammaticalization (pp. 107–159). Amsterdam/Philadelphia:
John Benjamins. doi: 10.1075/tsl.43.05giv
Harris, A. C. (2003). Preverbs and their origins in Georgian and Udi. In G. Booij, & J. van Marle
(Eds.), Yearbook of morphology 2003 (pp. 61–78). New York: Kluwer Academic Publishers.
doi: 10.1007/978-1-4020-1513-7_4
Harris, A. C., & Campbell, L. (1995). Historical syntax in cross-linguistic perspective. Cambridge:
Cambridge University Press. doi: 10.1017/CBO9780511620553
Heine, B. (1993). Auxiliaries: Cognitive forces and grammaticalization. Oxford: Oxford University
Press.
Heine, B., & Kuteva, T. (2002). World lexicon of grammaticalization. Cambridge: Cambridge
University Press. doi: 10.1017/CBO9780511613463
Hieber, D. W. (2014, May). Degrees and dimensions of grammaticalization in Chitimacha preverbs.
Paper presented at the 17th Annual Workshop on American Indigenous Languages (WAIL),
Santa Barbara.
Hieber, D. W. (2016). Semantic alignment in Chitimacha. Manuscript submitted for publication.
Hopper, P. J., & ompson, S. A. (1980). Transitivity in grammar and discourse. Language, 56(2),
251–299. doi: 10.2307/413757
Hopper, P. J., & Traugott, E. C. (2003). Grammaticalization (2nd ed.). Cambridge: Cambridge
University Press. doi: 10.1017/CBO9781139165525
Itkonen, E. (2005). Analogy as structure and process: Approaches in linguistics, cognitive psychology
and philosophy of science. Amsterdam/Philadelphia: John Benjamins. doi: 10.1075/hcp.14.
König, E., & Gast, V. (2008). Reciprocity and reexivity– description, typology, and theory. In
E.König, & V. Gast (Eds.), Reciprocals and reexives: eoretical and typological explorations
(pp. 1–32). Berlin/New York: De Gruyter Mouton.
Kuryłowicz, J. (1964). e inectional categories of Indo-European. Heidelberg: Carl Winter.
Lehmann, C. (2015). oughts on grammaticalization (3rd ed.). Berlin: Language Science Press.
© 2018. John Benjamins Publishing Company
All rights reserved
Chapter2. Category genesis in Chitimacha 
Los, B., Blom, C., Booij, G., Elenbaas, M., & van Kemenade, A. (2012). Morphosyntactic change:
A comparative study of particles and prexes. Cambridge: Cambridge University Press.
doi: 10.1017/CBO9780511998447
Martin, J. B. (2000). Creek voice: Beyond valency. In R. M. W. Dixon, & A. Y. Aikhenvald
(Eds.), Changing valency: Case studies in transitivity (pp. 375–403). Cambridge: Cambridge
University Press. doi: 10.1017/CBO9780511627750.013
Matthews, P. H. (2014). e concise Oxford dictionary of linguistics. Oxford: Oxford University
Press.
Mithun, Marianne. (1984). e evolution of noun incorporation. Language, 60(4), 847–894.
doi: 10.1353/lan.1984.0038
Roberts, I., & Roussou, A. (2003). Syntactic change: A minimalist approach to grammaticalization.
Cambridge: Cambridge University Press. doi: 10.1017/CBO9780511486326
Schladt, M. (2012). e typology and grammaticalization of reexives. In Z. Frajzyngier, & T. S.
Curl (Eds.), Reexives: Forms and functions (Vol. 1, pp. 103–124). Amsterdam/Philadelphia:
John Benjamins. doi: 10.1075/tsl.40.05sch
Swadesh, M. (1939a). Chitimacha grammar, texts and vocabulary. American Council of Learned
Societies Committee on Native American Languages (Ms. 497.3 B63c G6.5). Philadelphia:
American Philosophical Society Library.
Swadesh, M. (1939b). Chitimacha texts. In M. Swadesh, Chitimacha grammar, texts and vocab-
ulary. American Council of Learned Societies Committee on Native American Languages
(Ms. 497.3.B63c G6.5). Philadelphia: American Philosophical Society Library.
Swadesh, M. (1939c). Chitimacha-English dictionary (copy 1). In M. Swadesh, Chitimacha gram-
mar, texts and vocabulary. American Council of Learned Societies Committee on Native
American Languages (Ms. 497.3.B63c G6.5). Philadelphia: American Philosophical Society
Li brar y.
Swadesh, M. (1939d). Chitimacha grammar (copy 1). In M. Swadesh, Chitimacha grammar, texts
and vocabulary. American Council of Learned Societies Committee on Native American
Languages (Ms. 497.3.B63c G6.5). Philadelphia: American Philosophical Society Library.
Swadesh, M. (1946). Chitimacha. In H. Hoijer (Ed.), Linguistic structures of Native America
(pp.312–336). New York: Viking Fund.
Swanton, J. R. (1920). A sketch of the Chitimacha language [Unpublished grammar sketch].
National Anthropological Archives (Ms. 4122). Suitland: Smithsonian Institution.
Traugott, E. C., & Trousdale, G. (2010). Gradience, gradualness, and grammaticalization: How do
they intersect? In E. C. Traugott, & G. Trousdale (Eds.), Gradience, gradualness, and gram-
maticalization (pp. 19–44). Amsterdam/Philadelphia: John Benjamins. doi: 10.1075/tsl.90.
Traugott, E. C., & Trousdale, G. (2013). Constructionalization and constructional changes. Oxford:
Oxford University Press. doi: 10.1093/acprof:oso/9780199679898.001.0001
Tuggy, D. (2007). Schematicity. In D. Geeraerts, & H. Cuyckens (Eds.), e Oxford handbook of
cognitive linguistics (pp. 82–116). Oxford: Oxford University Press.
Watkins, C. (1964). Preliminaries to the reconstruction of Indo-European sentence structure. In
H. G. Lunt (Ed.), Proceedings of the 9th International Congress of Linguists (pp. 1035–1044).
e Hague: De Gruyter Mouton.
© 2018. John Benjamins Publishing Company
All rights reserved
 Daniel W. Hieber
Appendix. Abbreviations
1 rst person
2 second person
3 third person
agent
 adreditive
 andative
 auxiliary
 adjectivizer
 causative
 circumlative
 conditional
 continuative
 copula
 debitive
 demonstrative
 determiner
  dislative
 distal
 detransitivizer
 ergative
 existential
 do by handling
 imperative
 inceptive
 inchoative
 interrogative
 intransitive
 imperfective
 irrealis
 locative
 necessitative
 negation
 non-rst person
 nominalizer
patient
 perfect
 plural
 pluractional
 postposition
 preverb
 proximate
 participle
 punctual
 quotative
 reciprocal
 reditive
 reexive
 repetitive
 responsive
 reversive
 subordinator
 singular
 sociative
 stative
 sublative
 subordinative
  super-lative
 superreditive
 temporal subordinator
 topic marker
 transitive
 translative
 transreditive
 venitive
 vertical position
... Nearly all the studies referenced in this chapter adopt the particularist approach to word classes. The typological approach to word classes is still fairly recent, and little research has looked at North American languages from a constructional perspective (though see Hieber [2018] and Vigus [2018]). However, since this chapter is a crosslinguistic survey, I adopt the typological approach here. ...
... For example, prepositional uses of the word absent (as in the utterance absent those ropes, we'd float to a new and faraway place [COCA]) arose only in the 1940s (Harper 2020). In North American languages, one somewhat common closed class of words is the preverb category, words which form a semantic unit with their verb, and often indicate things like direction or aspect (Los et al. 2012: Ch. 1).2 Chitimacha has a closed set of 10 preverbs, shown below in (6) (Hieber 2018). By contrast, Menominee (Algic) has a large open class of preverbs (Bloomfield 1962: 214 While open classes tend to be lexical ones and closed classes tend to be functional ones, this is just a tendency (Velupillai 2012: 115). ...
Preprint
Full-text available
This chapter is a survey of word classes in indigenous North American languages, with the aim of providing an introduction to the study of parts of speech, and of highlighting the unique place and contribution of North American indigenous languages in this research. Section 2 defines lexical vs. grammatical and open vs. closed classes, and how these distinctions are realized in North American languages. Section 3 summarizes the prominent themes in word classes research in North America: 1) at what level a word is categorized (root, stem, or inflected word), 2) whether a given language distinguishes noun and verb, and 3) whether a given language has an adjective category. The chapter concludes that North American languages present serious challenges to the definition and status of word classes in linguistic theory, and that the development of distinct lexical categories in a language is not necessarily a given.
... guage of the last several speakers (Duralde 1802;Gatschet 1881a;Swanton 1908;Swadesh 1930) and prepared drafts of descriptive grammars (Swanton 1920;Swadesh 1939a), a dictionary (Swadesh 1939b), and a collection of stories (Swadesh 1939c), those materials were archived and never published. As a result few published descriptions of Chitimacha grammar exist (Swadesh 1933;Swadesh 1934a;Swadesh 1946a;Iannucci 2009;Hieber 2018;Hieber 2019b). The Chitimacha tribe learned of the existence of the archival materials in the 1990s, sparking a vibrant revitalization movement, including daily language and culture classes at the tribal school, the production of Rosetta Stone language learning software, and ongoing work on a dictionary and pedagogical grammar. ...
... This software was released in 2010 and is now provided free to every tribal member and is incorporated into the language curriculum at the tribal elementary school. The recent availability of digital copies of archival materials has also facilitated a wave of new research on the language (Weinberg 2008;Iannucci 2009;Brown, Wichmann & Beck 2014;Hieber 2018Hieber , 2019aHieber , 2019b. The revitalization team continues to collaborate with the author on a modern dictionary, with eventual plans for a pedagogical grammar as well. ...
Preprint
Full-text available
Preprint of a chapter for The languages and linguistics of indigenous North America: A comprehensive guide, eds. Carmen Jany, Keren Rice, & Marianne Mithun.
... A recent proposal (though not widely accepted) also suggested a long-distance genetic relationship with Mesoamerican languages (Brown, Wichmann, and Beck 2014). Numerous studies have examined various aspects of Chitimacha grammar and diachrony, including the development of its class of preverbs (Hieber 2014a(Hieber , 2018, its system of verbal person marking and agent-patient alignment (Hieber forthcoming), the structure of Chitimacha discourse (Hieber 2016a), and verbal valency and transitivity (Hieber 2016b(Hieber , 2017. ...
Chapter
The history of the Chitimacha language is a remarkable story of cultural survival. This chapter tells a part of that story, discussing the interactions between Chitimacha and other languages in the Southeast prior to colonial contact, the persecution of the Chitimacha people under the French, the language’s documentation by early linguists and anthropologists, and finally its modern revitalization.
... A recent proposal (though not widely accepted) also suggested a long-distance genetic relationship with Mesoamerican languages (Brown, Wichmann, and Beck 2014). Numerous studies have examined various aspects of Chitimacha grammar and diachrony, including the development of its class of preverbs (Hieber 2014a(Hieber , 2018, its system of verbal person marking and agent-patient alignment (Hieber forthcoming), the structure of Chitimacha discourse (Hieber 2016a), and verbal valency and transitivity (Hieber 2016b(Hieber , 2017. ...
Chapter
Full-text available
The history of the Chitimacha language is a remarkable story of cultural survival. This chapter tells a part of that story, discussing the interactions between Chitimacha and other languages in the Southeast prior to colonial contact, the persecution of the Chitimacha people under the French, the language’s documentation by early linguists and anthropologists, and finally its modern revitalization.
Article
Full-text available
Middle markers are characterized by a distribution halfway between grammar and the lexicon: with some verbs, middle marking encodes valency change, while with others it obligatorily occurs with no obvious synchronic motivation. Despite the existing cross-linguistic work on middle markers, their history is still largely unknown. In the typological literature, the standard view is that middle markers predominantly have their origin in reflexive markers, and that, in their development, it is invariantly the grammatical component that expands to the lexical component. In this paper, I challenge these assumptions based on the analysis of a sample of 129 middle marking languages. As I show, the sources and pathways whereby middle markers come about are much more numerous and varied than what has been reported in the literature. By taking a source-oriented approach, I also discuss how recurrent cross-linguistic trends in the distribution of middle markers can in part be explained by looking at their history.
Article
Full-text available
The middle voice is a notoriously controversial typological notion. Building on previous work (e.g. Kemmer, Suzanne. 1993. The middle voice . Amsterdam & Philadelphia: John Benjamins), in this paper I propose a new working definition of middle markers as inherently polyfunctional constructions which are partly associated with valency change in opposition to bivalent (or more) verbs and partly lexically obligatory with monovalent verbs. Based on this definition, the paper undertakes a systematic survey of 149 middle voice constructions in a sample of 129 middle-marking languages. Evidence from the sample shows that middle voice systems display a much richer variation in forms and functions than is reported in the literature. This richer empirical evidence challenges some of the mainstream views on middle marking, especially its purported connection with reflexivity and grooming-type events, and calls for an overall rethinking of the typology of the middle voice.
Article
Full-text available
This paper describes the alignment system for verbal person-marking in Chitimacha, a language isolate of Louisiana. Using data from recently digitized versions of texts collected by Morris Swadesh in the 1930s, I show that Chitimacha exhibits a split alignment system with agent-patient alignment in the first person and nominative-accusative alignment in non-first persons. The agent-patient alternation is shown to cross-cut subjects of intransitives, objects and even subjects of transitives, and direct/indirect objects of di-transitives. The agent-patient system in Chitimacha is therefore sensitive not to transitivity but rather to the semantic categories of agent and patient, making it an exemplary case of semantic alignment. I also discuss evidence of the diachronic origins of the agent-patient pattern and show that it arose via a reanalysis of transitive verbs with impersonal subjects ("transimpersonals") as intransitive verbs with patientive subjects.
Chapter
This book provides a general perspective on valency-changing mechanisms - passives, antipassives, causatives, applicatives - in the languages of the world. It contains a comprehensive typology of causatives by R. M. W. Dixon, and detailed descriptions of valency-changing mechanisms in ten individual languages by leading scholars, based on original fieldwork. The sample languages span five continents and every kind of structural profile. Each contributor draws out the theoretical status and implications of valency-changing derivations in their language of study, and the relevant parameters are drawn together, and typological possibilities delineated, in the editors' introduction. The volume, originally published in 2000, will interest typologists, those working in the fields of morphosyntactic variation and lexical semantics, and exponents of formal theories engaging with the range of linguistic diversity found in natural language.
Chapter
Book
Aims and Scope This collection of original papers is a representative survey of recent theoretical and cross-linguistic work on reciprocity and reflexivity. Its most remarkable feature isits combination offormal approaches, case studies on individual languages and broad typological surveys in one volume, showing that the interaction of formal approaches to grammar and typology may lead to new insights and results for both fields. Among the major issues addressed in this volume are the following: How can our current knowledge about the space and limits of variation in the relevant domain be captured in a structural typology of reciprocity? What light can such a typology shed on the facts of particular languages or groups of languages (e.g. Austronesian)? How can recent descriptive and typological insights be incorporated into a revised and more adequate version of the Binding Theory? How do verbal semantics, argument structure and reciprocal markers interact? How can we explain the pervasive patterns of ambiguity observable in these two domains, especially the use of the same forms both as reflexive and reciprocal markers? What are the major sources in the historical development of reciprocal markers? This combination of large-scale typological surveys with in-depth studies of particular languages provides new answers to old questions and raises important new questions for future research. © 2008 by Walter de Gruyter GmbH & Co. KG, D-10785 Berlin. All rights reserved.
Book
All languages have demonstratives, but their form, meaning and use vary tremendously across the languages of the world. This book presents the first large-scale analysis of demonstratives from a cross-linguistic and diachronic perspective. It is based on a representative sample of 85 languages. The first part of the book analyzes demonstratives from a synchronic point of view, examining their morphological structures, semantic features, syntactic functions, and pragmatic uses in spoken and written discourse. The second part concentrates on diachronic issues, in particular on the development of demonstratives into grammatical markers. Across languages demonstratives provide a frequent historical source for definite articles, relative and third person pronouns, nonverbal copulas, sentence connectives, directional preverbs, focus markers, expletives, and many other grammatical markers. The book describes the different mechanisms by which demonstratives grammaticalize and argues that the evolution of grammatical markers from demonstratives is crucially distinct from other cases of grammaticalization.