LINGUIST List 12.810

Fri Mar 23 2001

Review: Morse, Cubeo grammar; Miller, Desano grammar

Editor for this issue: Terence Langendoen <>

What follows is another discussion note contributed to our Book Discussion Forum. We expect these discussions to be informal and interactive; and the author of the book discussed is cordially invited to join in. If you are interested in leading a book discussion, look for books announced on LINGUIST as "available for discussion." (This means that the publisher has sent us a review copy.) Then contact Simin Karimi at or Terry Langendoen at


  1. Michael Cysouw, reviews Morse/Maxwell (1999) and Miller (1999)

Message 1: reviews Morse/Maxwell (1999) and Miller (1999)

Date: Fri, 23 Mar 2001 13:00:21 +0100
From: Michael Cysouw <>
Subject: reviews Morse/Maxwell (1999) and Miller (1999)

Morse, Nancy L. and Michael B. Maxwell (1999) Cubeo Grammar, Studies
in the Languages of Colombia 5 (SIL Publication in Linguistics; 130)
Summer Institute of Linguistics, Arlington. (X + 197 pp., $29.00,
ISBN 1-55671-044-5)

Miller, Marion (1999) Desano Grammar, Studies in the Languages of
Colombia 6 (SIL Publication in Linguistics; 132) Summer Institute of
Linguistics, Arlington. (X + 178 pp., $25.00, ISBN 1-55671-076-3)

reviewed by Michael Cysouw, ZAS Berlin

1. Introduction

The two books under review are both grammars of Tucanoan languages, a
small linguistic family in the Vaupes, located on the border of
Colombia and Brazil in South America. Until very recently, hardly
anything had been published about this linguistic family (cf. Kok,
1922; Espinoza Peres, 1955; Wheeler, 1967). This situation has
changed drastically in the last decade by a series of publications
under the auspices of the Summer Institute of Linguistics: two short
sketches with basic materials (Gralow, 1993; Smothermon et al.,
1995), some articles on select topics and two short, yet full fledged
grammars in the series 'Studies in the Languages of Colombia' (Jones
& Jones, 1991; Strom, 1992). And now there are these two grammars in
this same series, one on Cubeo by Morse & Maxwell and one on Desano
by Miller.

Section 2 of this review gives a synopsis of the two books with some
general comments. The books will then be put to the cross-linguistic
test: I will hold them in the light of the world's linguistic
diversity and differentiate those aspects that are unusual
cross-linguistically from those that are to be found often in the
world's languages. Section 3 highlights some cross-linguistically
unusual, but typical Tucanoan features. In contrast, section 4
discusses these languages with respect to some recent typological
claims. Section 5 concludes this review.

2. Synopsis

The Cubeo grammar consists of 6 chapters and two appendices. The
first chapter is a short description of phonological properties.
Chapter 2 is a long and extensive discussion of the verbal morphology
and chapter 3 is a thorough discussion of the nominal morphology.
Chapter 4 discusses other word classes, like adjectives, adverbs an
postpositions. Chapter 5 is a rather short note on clausal structure,
followed in chapter 6 by a lengthy discussion of subordination (which
subsumes relative and adverbial clause formation). The first appendix
is very useful; it presents an alphabetic list of all affixes - of
which there are many. A short notice on practical orthography
concludes the book.

In general, the Cubeo grammar has a clear structure and very
accessible explanations of the intricacies of the language. A
discussion of discourse devices is left out - with reference to an
article by Salser & Salser (1979). It is a pity that, for
completeness sake, the authors did not add a summary of that paper to
their description. I only dare to make one little addition to the
numerous fine morphological analyses in this grammar, concerning the
verb 'kiwA' (the /i/ should have a dash through it), meaning 'to
have'. Morse & Maxwell (58) note that this root is noncompositional.
However, it is composed of a locative copula and a causative suffix,
which results in a clear compositional meaning 'causing to be at' - a
cross-linguistic common expression of possession.

The Desano grammar consists of 11 chapters. The first chapter
presents a quick survey of the constituent order and of the
phonology. Chapter 2 is a short introduction of the word classes.
Chapter 3 and 4 discuss noun phrase structure; chapter 5 and 6
discuss verb phrase structure. Chapter 7 is a short note on sentence
structure. Then follow two thematic chapters, chapter 8 on question
formation and chapter 9 on negation. Finally, chapter 10 discusses
subordination (which -again- subsumes relative and adverbial clause
formation) and chapter 11 discusses some pragmatic devices that
structure the discourse in Desano.

In general, the outline of the Desano grammar is somewhat messy. Many
topics are discussed in various places throughout the book, which
makes it sometimes difficult to find a particular point of interest.
Also, an appendix with a list of affixes (like in the Cubeo grammar)
is dearly missed. Yet, this book displays a special fondness of the
author towards lexicography, as can be seen from long lists of
nominal classifiers (Miller 35-40) and possibilities of verb
compounding (Miller 88-108), and from a reference to a Desano
dictionary that is in preparation by the same author. The detailed
information on lexical collocations is the kind of data that is not
often found in descriptive grammars.

Both grammars excel in exemplifying all arguments with ample
sentences from actual discourse. Only in incidental cases sentences
from bible translations are used to corroborate a presented analysis.
It is difficult to judge how 'native' these translations are, but the
long participation of the authors in the communities suggests that
their judgements are to be trusted. All examples are given as
phonemic transcription, which makes them much easier to read, yet it
makes the data rather useless for phonologists. In particular, data
on interesting characteristics like nasal spreading (see below) can
not be gathered from these books.

3. Features of special interest

The Tucanoan languages are almost exclusively suffixational. They use
suffixes elaborately in an agglutinating fashion. The grammar of
Cubeo lists more than 100 monosyllabic suffixes and more than 70
polysyllabic ones (the situation is comparable in Desano). Note that
the language has only a small phonological inventory (six vowels, 10
consonants and phonemic nasalisation) and a very simple (C)V syllabic
structure, which results in only 132 theoretically possible
syllables. The chance that two of the 100 different suffixes are
accidentally homophonous is thus rather large. And indeed, many
suffixes have multiple meaning with no clear historical or semantic

A very interesting feature of the Tucanoan languages is nasal
spreading. Both Cubeo and Desano have six phonemic vowels, all of
which can be phonemically nasal or non-nasal. Yet, some non-root
vowels are unspecified for nasality. These vowels can become
nasalised under influence of inherently nasalised vowels:
'Nasalisation spreads from a nasal vowel to the right across morpheme
boundaries onto vowels which are unmarked for nasality, providing
only vowels unmarked for nasality and nasalizable consonants
intervene. ... Nasal spreading is blocked by an inherently oral
suffix. ... Nasal spreading is blocked by word boundaries' (Morse &
Maxwell 8). Contra Kaye (1971), both grammars argue for productive
spreading of nasalisation from left to right only. Spreading from
right to left occurs only in a small closed set of constellations.

Another cross-linguistically unusual, but typical characteristic of
the Tucanoan languages is the regular marking of evidentiality (cf.
Barnes, 1984). Each finite verb obligatorily marks how the speaker
obtained his information. Desano and Cubeo both distinguish four
evidential possibilities (Miller 64-68; Morse & Maxwell 32-36). The
first three are identical, although different names are used:
'visual/witnessed' (the speaker himself experienced the event),
'hearsay/reportative' (the speaker obtained his information from
someone else) and 'assumed' (the speaker supposes that an event has
occurred based on his knowledge of how things work).The description
of the fourth evidential differs. In Desano it is explained as
'inferred' (the speaker makes an inference) and in Cubeo as
'probable' (the speaker considers an event probable, based on his

Finally, a typical Tucanoan characteristic is a curious kind of
verb-subject agreement. The verb agrees with in gender with third
person animate subjects; typically, there is a three-way distinction
masculine-feminine-plural. The curious aspect of this agreement
system lies in the fourth available agreement morpheme, which is used
for agreement with inanimates and also for agreement with first and
second person. This combination is clear counterexample to the
animacy-hierarchy (Miller 64 ff.; Morse & Maxwell 38 ff.).

4. Typological considerations

The classic issue of linguistic typology is the cross category
harmony of word/constituent order (e.g. Hawkins, 1983).The basic
characteristic in this context is the order of verb and object in the
main declarative sentence. Both Cubeo and Desano are prominently OV,
although both allow VO as well. Counts of actual texts show a
preference for OV for both languages. As for the order of subject and
verb, there is less consistency. Both SV and VS are possible; for
Cubeo, the authors claim a preference for VS (leading to OVS as the
unmarked order - a rather rare case), but for Desano the expected SV
order is more prominent, although there are many cases of VS as well
(Miller 2, Morse & Maxwell 141). In all further aspects, both Cubeo
and Desano are typical examples of OV languages: they have
postpositions; the auxiliary follows the verb (V-Aux); the relative
clause follows the noun (N-Rel); the possessor precedes the possessee
(Gen-N); the adjective precedes the noun (Adj-N) and the determiner
precedes the noun (Det-N). Even in morphological aspects, these
languages behave according to the typological OV-'blueprint': they
are almost exclusively suffixational and incorporated nouns precede
the verb. There are only a few deviations from the general pattern -
mainly concerning adjectives, which are cross-linguistically often
the odd one out. In Cubeo, the adjective and the demonstrative can be
placed after the noun, adding special emphasis (Morse & Maxwell
91-92) and in Desano, the adjective is placed after the noun if the
noun has also a demonstrative and/or a numeral modifier (Miller 3).

A more recent result of typological comparison is Stassen's (1997)
investigation of intransitive predication. Among many other
cross-linguistic claims, Stassen notes a correlation between the fact
that predicatively used adjectives are treated like verbs and the
fact that verbs are not obligatorily marked for a distinction between
present and past. For both Cubeo and Desano this claim roughly holds.
Almost all predicative adjectives are treated like verbs (Miller
24-25, 51; Morse & Maxwell 123-124), only a small closed class of
nominal adjectives is found - including age terms ('old, new') and
dimensional terms ('small, large'). At the other side of the
correlation, Cubeo is clearly non-tensed as the marking of
present/past is only one of the various tense-aspect possibilities
(Morse & Maxwell 38 ff.). The tensedness type of Desano is less
clear. Verbs in the assumed evidential are obligatorily marked for
present/past (and are thus tensed, following Stassen's definition),
but for all other verb forms, the present/past distinction is only
one of various tense-aspect possibilities (and are thus non-tensed).
As the larger part of verbal forms are not marked for tense, Desano
roughly qualifies for Stassen's claim (Miller 64 ff.). Confusingly,
Stassen included two other Tucanoan languages in the sample of his
1997 study (Barasano and Retuara) but he classified them rather
differently from my classification of Cubeo and Desano. A more
careful investigation of the Tucanoan family in this respect seems

Another recent typological comparison at hand is Haspelmath's (1997)
investigation of indefinite pronouns. Both the Cubeo and the Desano
grammar are silent about this issue, in accord with Haspelmath's
lament that indefinite pronouns are often neglected in grammatical
descriptions. On Cubeo, Morse and Maxwell (124-125) only have a short
note on indefinite adjectives, but nothing on indefinite pronouns. On
Desano, Miller (138) only notes that inherently negative quantifiers
do not exist. Instead of the expression 'no one helped me' the
language uses a phrase with the literal translation of 'the one to
help me doesn't exist'. This might point towards a language that does
not have indefinite pronouns at all (a possibility offered by
Haspelmath, 1997: 56-57), though more information is necessary to
decide on the type of these Tucanoan languages in the typology of
indefinite pronouns.

5. Conclusion

Both books do a great job in giving an outline of the grammatical
structure of each language in less than 200 (small) pages - including
numerous examples and meticulous descriptions of the meaning and
usage of the intricate morphological structure. Judging from the ease
with which the authors deal with the languages, there is much more
knowledge on the languages backing these rather short descriptions. I
hope that the authors continue to share that hard-earned knowledge
with the wider linguistic community in the form of more detailed
publications on these fascinating languages.

6. References

- Barnes, Janet (1984). 'Evidentials in the Tuyuca Verb'.
International Journal of American Linguistics 50 (3): 255-271.
- Espinoza Peres, Lucas (1955). Contribuciones Linguisticas y
Etnograficas sobre algunos pueblos indigenas del amazonas Peruano.
Madrid: consejo superior de investigaciones cientificas.
- Gralow, Frances L. (1993). Un Bosquejo del Idioma Koreguaje.
Santaf� de Bogot�: Instituto Ling��stico de Verano.
- Haspelmath, Martin (1997). Indefinite Pronouns. (Oxford Studies in
Typology and Linguistic Theory). Oxford: Clarendon Press.
- Hawkins, John A. (1983). Word order universals. New York: Academic Press.
- Jones, Wendell & Paula Jones (1991). Barasano Syntax. (Studies in
the Languages of Colombia; 2). Arlington: Summer Institute of
- Kaye, Jonathan Derek (1971). 'Nasal harmony in Desano'. Linguistic
Inquiry 2: 23-56.
- Kok, P.P. (1922). 'Ensayo de gram�tica Dagseje o Tokano'. Anthropos
17: 838-865.
- Salser, J.K. & Neva Salser (1979). 'Some features of Cubeo
Discourse and Sentence Structure'. In: Robert E. Longacre & Frances
Woods (eds.) Discourse Grammar: studies in indigenous languages of
Columbia, Panama and Ecuador. Vol. 2, pp. 253-272. (Summer Institute
of Linguistics Publications in Linguistics and Related Fields; 52).
Arlington: Summer Institute of Linguistics.
- Smothermon, Jeffrey R., Josephine H. Smothermon & Paul S. Frank
(1995). Bosquejo del Macuna: Aspectos de al cultura material de los
macunas. Santaf� de Bogot�: Instituto Ling��stico de Verano.
- Stassen, Leon (1997). Intransitive Predication. (Oxford Studies in
Typology and Linguistic Theory). Oxford: Clarendon Press.
- Strom, Clay (1992). Retuar� Syntax. (Studies in the Languages of
Colombia; 3). Arlington: Summer Institute of Linguistics.
- Wheeler, Alva (1967). 'Grammatical Structure in Siona Discourse'.
Lingua 19: 60-77.

7. About the author

The author's research deals mainly with world-wide cross-linguistic
comparison of morphosyntactic characteristic. Currently, I am
investigating pronominal clitics from a typological view. I am
affiliated to the Zentrum fuer Allgemeine Sprachwissenschaft (ZAS),
Jaegerstrasse 10/11, 10117 Berlin, Germany, reachable at
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue