ChapterPDF Available

Valency classes in Yucatec Maya



Full description of valency classes of Yucatec Maya in the framework of the Leipzig valency database
Christiani Lehmanni inedita, publicanda, publicata
Valency classes in Yucatec Maya
huius textus situs retis mundialis
dies manuscripti postremum modificati
occasio orationis habitae
Conference on Valency Classes in the World's Languages.
Max Planck Institute for Evolutionary Anthropology,
Leipzig, April 14-17th, 2011
volumen publicationem continens
Comrie, Bernard & Malchukov, Andrej (eds.), Valency
classes. A comparative handbook. Berlin & New York:
Mouton de Gruyter
annus publicationis
Valency classes in Yucatec Maya
Christian Lehmann
1 Introduction
Yucatec Maya, called Maya by its speakers, is the Mayan language spoken on the peninsula of
Yucatan, in the east of Mexico. Together with Lacandón, Mopán and Itzá, it constitutes the
Yucatecan branch of the Mayan languages. There are some 850,000 ethnic Mayas, more than half
of whom have Maya as their first language and use it for daily communication. Some people over
70 years of age are still monolingual. The rest of the population has Spanish as their first language,
and its percentage is rapidly increasing. The language can be used, and is occasionally used, for
written communication by a small fraction of the population, but in general written communication
takes place in Spanish.
The language has been described in grammars and dictionaries since the 16
century. However,
no syntactic description is available.
Most of the data of the present study were gathered during two field-trips in 2011 and 2012.
Examples lacking a source indication stem from these two field trips.
The source is identified for
examples drawn from published texts.
2 Some elements of Yucatec grammar
Yucatec possesses the following major word classes: substantive, adjective, verb, verboid, adverb,
preposition. Minor word classes include auxiliary, numeral, numeral classifier, possessive classifier,
quantifier, pronoun, determiner, particle, conjunction. Adjectives are noun-like. There is little
nominal and much verbal morphology. Yucatec morphology comprises compounding,
incorporation, derivation and inflection and is mildly synthetic. Word order is essentially right-
branching, with the exception of determiners and adjective attributes, which obligatorily and
optionally, respectively, precede the nominal head.
2.1 Verbal clause
The clause core starts with the predicate. If it is verbal, it may be followed by complements; then
comes the subject, and adjuncts tend to follow the subject. A simple verbal sentence (S) is
introduced by a non-inflecting – auxiliary (Aux) or modal verboid (see §4.4) which fixes the
aspectual or modal category of the sentence. Some auxiliaries trigger a deictic clitic (DC) at the end
of the clause. The structure of the clause is, thus, as shown in S1 and illustrated by E1.
I thank Henrike Frye B.A. for doing part of the fieldwork in 2011, and Gaspar Maglah Canul from Kantunilkin and
Ernesto May Balam and Ramón May Cupul from Yaxley, Quintana Roo, for contributing data and insights on Yucatec.
Christian Lehmann, Valency classes in Yucatec Maya 2
S1. Verbal clause
[ [ X ]
[ Y ]
( [ -e' ]
) ]
E1 ts'o'k u beet-ik u kool in taatah
[ [ TERM ]Aux [ SBJ.3
milpa POSS.1.SG father
‘my father already did his cornfield’
Topicalization and focusing allow verbal dependents to be placed in front of the clause. Since many
of the example sentences are elicited translations, topicalization by left-dislocation is omnipresent
in them, starting with E9. The topic is not a constituent of the clause following it and may be
omitted without harm to its grammatical structure.
The auxiliary is the structural head of the autonomous verbal clause (S); the rest is the verbal
clause core (VCC), which suffices for many subordinate constructions. The verbal clause core
consists of the verbal complex (VC) and its dependents. All of the latter are syntactically optional
under all circumstances.
The verbal complex, in turn, centers around the finite verb (V.fin). It is
introduced by the pronominal clitic (PC) Pn under conditions to be specified presently, as shown in
S2 and illustrated by E2.
S2. Verbal complex
[ [ Pn ]
[ W ]
E2 u y-ah-s-ik-ech
[ [ SBJ.3 ]
[ 0-wake-CAUS-INCMPL-ABS.2.SG ]
‘(that) he wakes you up’
The finite verb, shown in S3 and illustrated by E3, contains a status suffix which codes modal and
aspectual categories and is conditioned syntactically – by the auxiliary in autonomous verbal
clauses, by the matrix construction in subordinate clauses.
S3. Finite verb
[ [ W ]
E3 ah-s -ik -ech
[ [ wake-CAUS ]
[ -2.SG ]
‘wake you up’
Apart from personal pronouns, there are two series of bound pronominal elements (Pn):
Pronominal clitics (PC) cross-reference the subject (S
) of the transitive verb, the subject of the
intransitive verb in incompletive status and, on the possessed nominal, its possessor (P
shown twice in E1).
Absolutive suffixes (A
) cross-reference the direct object of the transitive verb, the subject of
the intransitive verb in completive and subjunctive status and the subject of a non-verbal
The experiential construction with óol (S6 as instantiated by E8 and analogous constructions) is probably an exception
to this, as relevant predicates do not occur without this inner dependent. Given this, they might be analyzed as phrasal
Christian Lehmann, Valency classes in Yucatec Maya 3
Both of these sets have pronominal function. However, their referents may be represented, in
addition, by personal pronouns or noun phrases in the same clause; and in that case, the pronominal
elements agree with them in person and number. Yet, the absolutive suffix is mostly zero for third
person singular, and agreement with a third-person plural subject is optional, with the consequence
that these suffixes are seldom seen as agreeing with nominal dependents in their clause.
Formulating the distribution of the pronominal elements from the perspective of the finite verb:
this is preceded by a pronominal clitic and/or followed by a pronominal suffix according to the
following rules:
With transitive verbs, the pronominal clitic cross-references the subject, while the suffix cross-
references the direct object.
With intransitive verbs, the pronominal element cross-references the subject. The syntagmatic
slot chosen, with its paradigm, depends on the verb’s status, viz.: in completive and subjunctive
status, it is the suffix, in incompletive status, it is the clitic.
The alternation in the syntagmatic position of the Pn with intransitive verbs is hard to abstract over
in the construction formulas. In such constructions, the Pn will not be subscripted (implying that it
may alternatively be a suffix), while with transitive verbs, it will be subscripted by PC.
In accord with the syntactic functions that they correspond to, the pronominal clitics are called
subject or possessive clitics, while the pronominal suffixes are called absolutive suffixes. (In
traditional Mayan linguistics, they are called 'set A' and 'set B', resp.) However, these names are just
mnemonic for the set of syntactic functions that these morphemes cross-reference. Syntactically, the
language has pure accusative alignment.
The status suffixes display rich sets of allomorphs, chiefly conditioned by the transitivity of the
verb stem. Together with the two sets of pronominal elements, this guarantees that every verb form
occurring in a text is marked as either transitive or intransitive, with very few ambiguous cases.
Moreover, both of these principal valency classes are subdivided into conjugation classes, which,
again, take the form of allomorphy of status suffixes. Most of these conjugation classes are
productive. To the extent that they are the goal of derivational operations, some of the exponents of
conjugation classes allow an alternative analysis as derivational operators. One is familiar with such
a situation from SAE languages. In Latin, for example, the a-conjugation is the goal of the
verbalization that converts a noun like cumulus ‘pile’ into a verb like cumulare ‘pile up’, so that -a
may be regarded as a derivational operator. Two Yucatec cases will be discussed in sections 5.2.1
2.2 Possessive construction
The nominal possessive construction is as shown in S4 (see Lehmann 2002 for details). Its core is
the combination of the pronominal clitic (Pn) with the possessed nominal (Pd). While there is no
possessive construction without the Pn, the possessor (Pr) is optionally represented by a lexical NP.
If Pn and Pr are co-present, they agree in person and number (i). This is illustrated by E4.
S4. Nominal possessive construction
[ [ [ Pn
[ Pd ]
] ([ Pr
) ]
E4 u maamah le chaan xibpaal-o’
[ [ [POSS.3]
] [ DEM little man:child-D2 ]
‘the little boy’s mother’
Christian Lehmann, Valency classes in Yucatec Maya 4
As already said in §2.1, there is only one paradigm of pronominal clitics (PC). For the reader’s
convenience, the Pn of the construction formulas is glossed as S
if they introduce a verbal
construction, and as P
if they introduce a nominal construction.
2.3 Further remarks
The strategy of signaling syntactic relations is, thus, exclusively head-marking. There are no cases,
only prepositions. Apart from specific prepositions for local and other concrete functions, there is
an all-purpose preposition ti', glossed L
, which marks a generic local relation as well as the
indirect object.
The construction formulas which describe verbal valency frames and alternations represent
verbal clause cores and are therefore indexed with VCC. Since the auxiliary which lifts them to the
clause level does not matter for the analysis, it has been omitted. Furthermore, in all construction
formulas, it is understood that all main constituents except the predicate are optional. In order not to
clutter the formulas, the corresponding parentheses have been omitted. Constituents are represented
by capital letters serving as variables which are kept constant across formulas to the extent possible.
3 Nominal clause
3.1 Basic structure
The nominal predicate (W in S5) is a nominal or adjectival group. Just as the verbal predicate, it
precedes all other main constituents in its clause. It bears an absolutive suffix (-A
) by which it
agrees with any lexical or pronominal subject constituent (S) in person and number (i). However,
the agreement is seldom visible: If i = 3rd, the absolutive suffix is zero. If i = 1st or 2nd ps., S
only appears if it is emphatic. Only if i = 3rd is the agreement regularly visible. E5 illustrates
a substantival, E6 and E7 an adjectival predicate.
S5. Basic nominal predicate construction
[ [ W ]
[ S
E5 h-koolnáal-o’n
‘we are farmers’
E6 tikin le lu'm-o'
dry DEM earth-D2
‘the ground is dry’
S of S5 may be expanded into the possessive construction of S4. The resulting S6 is a subtype of
S5, where S is some body part sensu lato and E is an experiencer, as illustrated by E7. The index i is
necessarily third person (commonly singular), while E may vary through the persons and numbers.
S6. Adjectival experiential construction
[ [ W ]
[ P
[ E ]
Christian Lehmann, Valency classes in Yucatec Maya 5
E7 yah in k’ab
painful POSS.1.SG hand
‘my hand hurts’
A couple of experiential predicates, including sahak ‘afraid’ (E9), are ascribed to living beings
themselves. Most of them are, instead, ascribed to their óol, instantiating S in S6 and rendered by
‘mind’ in the interlinear gloss for want of a closer English counterpart.
Yah of E7 is, in fact, the
only predicate that may be ascribed both to óol and to other body parts. Óol is an inalienable noun
and takes the experiencer of an experiential construction as its possessor.
The construction is
exemplified by E8. Note that ok’om le chaan xch’úuppalo’ would be ungrammatical.
E8 ok'om u y-óol le chaan x-ch'úuppal-o'
sad POSS3 0-mind DEM little F-girl-D2
‘the little girl is sad’
Another example of S6 is E50 below.
3.2 Relational adjective construction
Some adjectives have valency, taking a complement via the preposition ti’. This construction is
formalized in S7 and illustrated in E9. S7 is an extension of S5 by the PrepP containing L.
S7. Relational adjectival predicate construction
[ [ W ]
[ ti’ [ L ]
[ S
E9 le wíinik-o' sahak ti' le báalam-o'
DEM man-D2 afraid LOC DEM jaguar-D2
‘The man is/was afraid of the jaguar.’
While the S of E9 is an experiencer and L the stimulus, there is no fixed association of semantic
roles with the constituents of S7. As an alternative, the stimulus may take the place of S, while L is
expanded into a possessive construction whose head (L) is some body part and whose possessor is
the experiencer (E). This expansion of S7 takes the form of S8 and is illustrated by E10.
S8. Relational adjectival experiential construction
[ [ W ]
[ ti’ [ P
[ E ]
[ S
E10 le xibpal-o' uts t-u y-ich u tumben baaxal
DEM boy-D2 good LOC-POSS.3 0-eye POSS.3 new toy
‘The boy likes his new toy.’
In analogy to the specification of S6, the body part noun L of S8 may be óol, as illustrated by E11.
It has a close semantic counterpart in the original [!] sense of German Mut as it appears in Großmut ‘magnanimity’,
Sanftmut ‘placidity’ or mir ist unbehaglich zumute ‘I feel uncomfortable’. However, even mosquitos have an óol, so
there is little room for psychological speculation on the Yucatec construction.
Details in Verhoeven 2007, ch. 8.1.1.
Christian Lehmann, Valency classes in Yucatec Maya 6
E11 yah t-in w-óol
painful LOC-POSS.1.SG 0-mind
u loobil-t-a'l le mehen paal-al-o'b-o’
SBJ.3 harm-TRR-PASS.INCMPL DEM small child-COLL-PL-D2
‘I feel sorry that the little children are maltreated.’
4 Verbal valency patterns
In terms of quantitative valency, Yucatec possesses monovalent, bivalent and trivalent verbs. The
latter are relatively few in number. There are no avalent verbs. The actant
of a monovalent verb is
its subject. Many monovalent verbs are impersonal, i.e. they take exclusively clausal subjects.
Bivalent verbs are overwhelmingly transitive. There are, however, some intransitive bivalent verbs,
taking a complement which may be represented by an adverbial or be introduced by a preposition.
Trivalent verbs take such a complement in addition to a subject and direct object.
4.1 Intransitive verb
4.1.1 Basic construction
The construction formula for all intransitive verbs is S9. Pn represents the pronominal element
cross-referencing the subject (S). It may be a pronominal clitic or an absolutive suffix as explained
in §2.1. There may be other verbal dependents in the construction, esp. local complements or
adjuncts, as in E14.
S9. Basic intransitive verb construction
[ [ Pn
] [ W ]
[ S
As said in §2.1, intransitive verbs differ from transitive ones by their conjugation classes. There are
three conjugation classes of intransitive verbs, which are rather closely associated with semantic
verb classes: active, inactive and fientive
intransitive verbs. Active and inactive verb stems differ
in the control of their subject. The three verb classes are, in their turn, illustrated by E12, E13 and
E12 le xch'úupal-o' h síit'-nah-ih
DEM girl-D2 PRFV jump-CMPL-ABS.3.SG
‘the girl jumped’
E13 h lúub le che'o'
PRFV fall(CMPL) DEM wood-D2
‘the tree fell’
The term actant will denote a nominal expression comprised by a verb’s structural valency, while the term argument
will denote a participant inherent in a verb’s meaning. The distinction will prove relevant, i.a., in §6.1.
In general, a fientive verb (Lat. fieri ‘become’) is a verb derived from an adjective and designating a change of state,
like Engl. to cool (down). Traditionally, such verbs were (somewhat inappropriately) called inchoative.
Christian Lehmann, Valency classes in Yucatec Maya 7
E14 le paal-al-o'b-o' h kul-chah-o'b ti' le k'áanche'-o'
‘the children sat down on the chair’
The three verb classes generate subtypes of S9, which contain V.intr_act, V.intr_inact and
V.intr_pos instead of the mere V.intr as the category label of W. These subtypes differ in their
derivational potential, as will be seen in §5.
Meteorological predicates like ‘rain’ are monovalent, as shown in E15:
E15 táan u k’áax-al ha’/chaak
PROG SBJ.3 rain-INCMPL water/rain
‘it is raining’
4.1.2 Intransitive experiential construction
The experiential construction with an inactive intransitive verb as predicate is as shown in S10. E
represents the experiencer, L the stimulus.
S10. Intransitive experiential verb construction
[ [ Pn
] [ W ]
[ P
[ E ]
[ ti’ [ L ]
E16 illustrates its most common appearance, with óol instantiating S. Just as experiential adjectives,
most experiential verbs only allow óol as their subject.
E16 háak’ in w-óol ti’ hun-túul ch’o’.
scare(CMPL.ABS.3.SG) POSS.1.SG 0-mind LOC one-CL.AN mouse
‘I got a fright because of a mouse’
S10 may be composed as follows: Take S7 as a point of departure; substitute an intransitive verb for
W; substitute the possessive phrase appearing in S6 for the S of S7.
4.1.3 Directed motion verb
Verbs of directed motion form a subclass of inactive verbs, although they do allow for a controlling
subject. They are such verbs as bin ‘go’, taal come’, máan ‘pass’, ook ‘enter’, hóok’ ‘exit’, eem
‘get down’, na’k ‘get up’ and a few more. They differ from other motion verbs – essentially manner
of motion verbs like xik’nal ‘fly’ – by allowing a local complement (L).
The category of this latter
is adverbial; and it may appear in the form of a prepositional phrase, introduced, more often than
not, by ti’. The overall construction is as shown in S11. It is similar to the relational adjectival
predicate construction (S7) and an expansion of the basic intransitive verb construction S9. E17 is
an example.
S11. Construction of intransitive verb with complement
[ [ Pn
] [ W ]
[ S
[ L ]
There are examples featuring the active intransitive manner-of-motion verb áalkab ‘run’ with a local dependent, which
may not be exceptional if the latter can be analyzed as an adjunct.
Christian Lehmann, Valency classes in Yucatec Maya 8
E17 le xibpal-o' h hóok' ti' u kaahal
DEM boy-D2 PRFV exit(CMPL) LOC POSS.3 village
‘the boy left his village’
Directed motion verbs are central to the verbal lexicon and important for the syntax. The first two
enumerated above have grammaticalized variants and also an irregular conjugation. Incidentally, as
a comparison among E14, E17, E22, E24 and E54 shows, prepositions do not distinguish among
local relations (essive, allative, ablative, perlative); these are coded as part of the verb meaning (cf.
Lehmann 1992).
4.1.4 Positional
In terms of verb derivational morphology, a fientive verb is an intransitive verb stem derived from a
non-verbal (generally, nominal) base, as schematized in S18 below. However, the suffix -tal, which
encodes this in Yucatec, is, at the same time, the exponent of one of the three intransitive
conjugation classes. That is, there is a class of verb roots that conjugate in that class; and moreover,
different verb statuses, aspects and moods fuse in that suffix, which phenomenon is more typical of
an inflection stem exponent than of a derivational operator. We will see a similar systematic
ambiguity in the extraversive suffix -t (§
The fientive conjugation class thus comprises roots and derived stems. Fientive verb roots are
comprised of positionals and a couple of other roots with related meanings. The set of positionals
comprises at least two dozen roots like wa’l ‘stand’, chil ‘lie’, xol ‘kneel’, t’uch ‘squat’ etc. Their
construction is the same as for directed motion verbs, viz. S11. Two positionals are illustrated in
E14 and E18.
E18 káa chil-lah-o'b wen-el hun-súutuk
CONJ lie-FIENT.CMPL-3.PL sleep-INCMPL one-moment
‘and they lay down to sleep a bit’ (ka'tuul_15)
The derivational potential of this verb class is further discussed in §5.2.1.
4.2 Transitive verb
4.2.1 Basic construction
The construction of a transitive verb is as formalized in S12 and illustrated by E19. A is the subject,
P is the direct object. Adjuncts are not shown.
S12. Basic transitive construction
[ [ Pn
[ W ]
[ P
[ A
E19 t-u méek'-ah u chaan xibpal le maamah-o'
PRFV-SBJ.3 hug-CMPL POSS.3 little boy DEM mother-D2
‘the mother hugged her little boy’
In synchronic grammar, S12 is a basic construction. Diachronically, it may be composed as follows:
Start from the basic nominal predicate construction S5. Expand its W to the possessive construction
S4. Replace the Pd nominal by a nominalized transitive verb. Thereby, its Pr gets interpreted as a
transitive subject, with its Pn functioning as subject clitic. The result is the transitive VCC S12.
Christian Lehmann, Valency classes in Yucatec Maya 9
Given the default-template character of S12 for two-participant situations, the semantic roles of
A and P are as variable as for the transitive verb in many SAE languages. For a transitive verb, the
direct object is obligatory. That does not mean that P must be overt. It means that j is referential; i.e.
it is identifiable in the linguistic or extralinguistic context. If that condition is not fulfilled, the verb
must be detransitivized by introversion; see §
There are two transitive conjugation classes which differ by having or lacking an exponent, viz.
a conjugation suffix directly after the stem, thus preceding the status suffix. Verb stems are assigned
to the two conjugation classes according to the classification of T1, to be read from left to right:
T1. Transitive conjugation classes
provenience complexity basicness phonotactics exponent
primitive transitive root 0 basic transitive root non-primitive transitive root 0
marked transitive root -t
complex stem (other than causative) -t
loan -t
Examples of these categories are:
primitive transitive root: pix ‘cover’, as in E20
non-primitive transitive root: méek’ ‘hug’, as in E19
marked transitive root: báats’ ‘smooth’ (not illustrated)
complex stem: háakchek’ ‘slide stepping on something’, as in E46
loan: formar ‘form’.
The -t suffix is identical to the extraversive operator, illustrated by tsikbal-t ‘tell’ in E23 and to be
discussed in § The subdivision of the basic transitive roots is irrelevant for the conjugation
class exponent, but the phonotactics underlying it conditions allomorphy in some verbal categories
to be discussed in §5.3.
4.2.2 More complex constructions
A semantically suitable subset of transitive verbs takes an instrumental adjunct. These include beet
‘make’, pa’ ‘break, smash’, kíins ‘kill’, hats’ ‘hit [with tool]’, xot ‘cut [with tool]’, pix ‘cover’, chup
‘fill’, huch’ ‘grind’ and koh ‘toss, touch’. The construction is represented as S13, which is an
expansion of the basic transitive construction S12 by the instrumental adjunct I. It is illustrated by
S13. Transitive verb with instrumental adjunct
[ [ Pn
[ W ]
[ P
[ A
[ yéetel [ I ]
E20 le x-ch'úup-o' t-u pix-ah le xibpal yéetel teep'el-o'
DEM F-woman-D2 PRFV-SBJ.3 cover-CMPL DEM boy with blanket-D2
‘the woman covered the boy with a blanket’
A marked transitive root is one that lacks a more basic intransitive counterpart but nevertheless requires the -t suffix
for inflection as if it were derived. This class is exceptional and has very few members.
Christian Lehmann, Valency classes in Yucatec Maya 10
For a better understanding of the derived experiential constructions to be discussed in §, a
curious gap in the grammar should be mentioned: although there are experiential constructions
involving óol on the basis of adjectives, intransitive and transitive verbs, and although there are
derived transitive verbs taking that noun as their object for which actor and experiencer are distinct,
there is no base transitive verb of the latter kind; all the transitive roots taking óol as their object
require a semantically reflexive constellation, as in E21.
E21 yéetel hun-p'éel libro u_ti'a'l a nay-ik a w-óol
and one-CL.INAN book in.order SBJ.2 amuse-INCMPL POSS.2 0-mind
‘and a book to amuse yourself’ (hnazario_104)
The expected variant with u nayik a wóol ‘which [book] may entertain you’ does not exist.
4.3 Trivalent verb
Trivalent verbs take a subject, a direct object and another complement which may be represented by
an adverbial or be introduced by a preposition. The animacy of the indirect complement conditions
the major subdivision here.
4.3.1 Trivalent verb with inanimate indirect complement
If the indirect complement is inanimate, it may appear in the form of an adverbial or of a
prepositional phrase introduced by a suitable preposition. The core of these trivalent verbs is formed
by directed transport verbs,
the transitive (semantically causative) counterpart to the directed
motion verbs seen in §4.1. The indirect complement then bears some local relation to the verb.
Directed transport verbs include ts’a’ ‘put’ and its local reverse ch’a’ ‘take [for oneself]’, the two
opposite deictic transport verbs bis ‘carry, take [to]’ and taas ‘bring’ (s. §, and furthermore
tul ‘push’, túuxt ‘send’, pul ‘throw’, t’oh ‘pour’, láal ‘pour’, but’ ‘fill’. The construction of these
verbs is shown in S14, where L is the indirect complement. E22 is an example.
S14. Construction of ditransitive verb with inanimate indirect complement
[ [ Pn
[ W ]
[ P
[ A
[ L ]
E22 le xibpal-o' t-u pul-ah le boolah teh béentaanah-o'
DEM boy-D3 PRFV-SBJ3 throw-CMPL DEM ball LOC:DEM window-D2
‘the boy threw the ball into the window’
S14 is the semantically causative counterpart to the construction of an intransitive verb with
complement (S11) and may therefore be composed by extending the basic transitive construction
(S12) by the same complement. Causative derivation of bivalent intransitive verbs is treated in
With some of the above verbs, the indirect complement may also be animate, and then the
construction is as in §4.3.2. The most important one among these is the basic trivalent verb ts’a’
‘put, give’. It has two constructions corresponding to its senses: With the meaning ‘put’, it belongs
An example of a non-directed transport verb is kuch ‘load [on oneself]’.
The two verbs bis and taas mentioned before are, in fact, causative derivations of bin ‘go’ and taal ‘come’, resp.; but
they are totally lexicalized and part of the basic vocabulary.
Christian Lehmann, Valency classes in Yucatec Maya 11
in the present class of directed transport verbs. With the meaning ‘give’, it belongs in the class of
verbs with an animate indirect complement.
4.3.2 Trivalent verb with animate indirect complement
The indirect complement of a trivalent verb may be typically or exclusively animate. The
construction is then a subtype of the generic construction of S14 formed by specifying the adverbial
L as a prepositional phrase and instantiating its preposition by ti’. The prepositional phrase may
then be categorized as an indirect object. The set of relevant verbs includes, first of all, ts’a’ ‘give’
(the same verb that figures in §4.3.1 with the meaning ‘put’), then communication verbs such as a’l
‘say’, k’áat ‘ask’, tsikbat ‘tell’, and furthermore e’s ‘show’, ta’k ‘hide’, k’am ‘get’, ookol ‘steal’.
The pattern is shown in S15 and illustrated by E23.
S15. Construction of ditransitive verb with animate indirect complement
[ [ Pn
[ W ]
[ P
[ A
[ ti’ [ L ]
E23 le xch'úupal-o' t-u tsikbal-t-ah hun-p'éel che'h-bil tsikbal
DEM girl-D2 PRFV-SBJ.3 talk-TRR-CMPL one-CL.INAN laugh-GER story
ti' le xibpal-o'
LOC DEM boy-D2
‘the girl told the boy a funny story’
The position of L depends on its lexical status: If it is a lexical NP, it takes the position indicated. If
it is a pronoun, the preposition is optional; and if it is missing, this indirect object immediately
follows the verb.
As indicated, the verb k’am ‘get’, with the goal of the transport in subject function, follows the
same pattern:
E24 t-in k'am-ah hun-p'éel kaartah ti' in maamah
PRFV-SBJ.1SG receive-CMPL one-CL.INAN letter LOC POSS.1SG mother
‘I received a letter from my mother.’
Cf. the remarks of §4.1.3 on the indifference of prepositions to local relations.
4.4 Verboids
Verboids share their valency frames with verbs, differing from these only by failing to inflect for
status or combine with the initial auxiliary of S1. Intransitive verboids include yaan ‘exist’, k’abéet
‘necessary’, taak ‘want’ and a few others. Transitive verboids include k’ahóol ‘know’, yáamah
‘love’, p’éek ‘hate’, k’áat ‘want’ and a few others. Verboids differ from most verbs by their stative
meaning. Intransitive verboids take absolutive suffixes (if they are personal at all) like intransitive
(completive or subjunctive) verbs; transitive verboids take pronominal clitics and absolutive
suffixes as transitive verbs do. Verboids follow the same construction patterns as verbs; for
instance, the construction of k’abéet in E25 (from E31.b below) is a variant of the intransitive verb
construction with complement (S11); the construction of k’ahóol in E26 is the basic transitive verb
construction (S12).
Christian Lehmann, Valency classes in Yucatec Maya 12
E25 k'abéet-ech teen
necessary-ABS.2.SG me
‘I need you’ (HK'AN_452.1)
E26 le xibpal-o' u k'ahóol le xch'úupal-o'
DEM boy-D2 SBJ.3 know DEM girl-D2
‘the boy knows/knew the girl’
Intransitive verboids are verbalized by the fientive derivation (§5.2.1), transitive verboids are
verbalized by extraversive verbalization (§
5 Verb-coded alternations
Yucatec has a number of derivational operations that change the category of the base, transferring it
into a different word class or just changing its relationality. These operations will be described not
as undirected alternations between two constructions of the same level, but as derivations or
transformations that derive a target construction from a base construction. The criterion of the
directionality is formal complexity: If constructions A and B are paradigmatically related and share
a common base, but B comprises an additional formative, then B is based on A rather than vice-
This section is not an exhaustive survey of Yucatec verb derivation (see Lehmann 1993 for a
more comprehensive account). With a few exceptions, only such operations will be considered as
change the relationality of the base. §5.1 briefly considers the deverbal derivation of adjectives,
which for transitive bases amounts to a valency reduction. §§5.2 and 5.3 treat the formation of
intransitive and transitive verbs in parallel fashion, to the extent appropriate.
5.1 Formation of adjectives
Here only deverbal adjectives will be considered. There is, in fact, a rich set of deverbal adjective
derivations, only the two most productive and regular of which will be reviewed.
5.1.1 Resultative adjective
S16 is a subtype of the adjectival predicate construction formalized in S5.
S16. Resultative adjective construction
[ [ [ W ]
-a’n ]
[ S
All transitive stems, as in E27.a, and all positional stems, as in #b, undergo the process. Resultatives
from other intransitive verbs occur sporadically.
E27 a. ts'ik-a'n Peedroh
shave-RSLTV Peter
‘Peter is shaved (i.e. in the state resulting from shaving)’
b. kul-a'n Hwaan
sit-RSLTV John
‘John is at home (lit.: seated)’
Christian Lehmann, Valency classes in Yucatec Maya 13
While S16 embeds the derived adjective into a clause structure, such adjectives may also be used
attributively. For transitive bases, the resultative derivation involves functional passivization, since
such an adjective modifies the undergoer, while the underlying actor is demoted to the typical
adjunct function associated with the passive (§ E28 illustrates a resultative adjective with
such an adjunct.
E28 chéen méek'-a'n bin tuméen h-p'óokinah tsuuk
just hug-RSLTV QUOT by M-hat:USAT:INTROV paunch
‘embraced, however, by Paunchhat’ (HK'AN_302)
5.1.2 Stative positional
Given a positional base as illustrated in E14, a stative positional may be derived. It has the structure
shown in S17 and illustrated by E29.
S17. Stative positional derivation
[ [ [ W ]
-Vkbal ]
[ S
E29 le paal-al-o'b-o' kul-ukbal-o'b ti' lu'm
DEM child-COLL-PL-D2 sit-POS-PL LOC earth
‘the children are/were sitting on the floor’
The suffix (glossed by -P
‘position’) starts with a vowel (V) that is subject to full harmony with
the root vowel. Describing this adjectival derivation in a clause formula such as S17, an
instantiation of S5, is appropriate as these adjectives are not used in attributive function.
For a positional base, there is no difference in meaning between the two adjectival derivations
in -a’n and -Vkbal.
5.2 Formation of intransitive verbs
A couple of derivations of intransitive verbs from intransitive bases will not be considered as they
do not affect valency. Instead, this section is subdivided as follows: After the formation of
intransitive verbs on nominal bases, which does not alter the valency, their formation from
transitive bases, i.e. by valency reduction, will be considered. This, in turn, is subdivided according
to the actant whose slot is blocked: first operations blocking the direct object slot, then operations
blocking or demoting the subject slot are considered. All of the derived constructions introduced in
this section are subtypes of the basic intransitive verb construction S9 or some extension of it.
5.2.1 Intransitive verbs from nominal bases: fientive verbs
Given the basic nominal predicate construction of S5, one may verbalize the nominal predicate by
applying the fientive derivation, schematized in S18, a subtype of S9. The base may be an adjective,
as in E30.b, or a noun of suitable meaning, as in #a. The operation transfers the stem into the
fientive conjugation class of intransitive verbs (see §4.1).
S18. Fientive derivation
[ [ Pn
] [ [ W ]
-tal ]
[ S
Christian Lehmann, Valency classes in Yucatec Maya 14
E30 a. ts’o’k u hach y-áak’ab-tal
TERM SBJ.3 really 0-night-FIENT.INCMPL
‘it has become night / is already night’
b. táan u tikin-tal le lu'm-o'
‘the ground is getting dry’
E30.b is the fientive counterpart to basic E6. The derivation does not affect the valency, but only
verbalizes the base, which may become necessary if the clause is to be marked for aspect or mood.
Depending on the aspect chosen, the aktionsart may also be more dynamic than with the stative
base, as shown by E30.b.
This operation applies in a completely regular fashion to all adjectival bases, conserving their
valency. Thus, among others, all the adjectives treated in §3 form a fientive, so that their nominal
experiential constructions have a verbal counterpart. Thus, beside yah ‘painful’, we find yahtal
‘hurt, ache’, and beside sahak ‘afraid’, there is sahaktal ‘fear, get frightened’. Likewise, the
operation applies to intransitive verboids like k’abéet ‘necessary’. E31.b shows the base form, #a
the verbalized form of this lexeme. The sentences are, in the order #a #b, utterances of the two
main characters of the tale. They show that the verboid and its verbalization may be synonymous.
E31 a. le k'iin k-in k'abéet-tal ti' teech-e'
DEM sun/day IMPF-SBJ.1.SG necessary-FIENT.INCMPL LOC you-TOP
káa taal-ak-ech a ch'a'-en
CONJ come-SUBJ-ABS.2.SG SBJ.2 take(SUBJ)-ABS.2
‘the day that you need me, you come to fetch me’ (HK’AN_198)
b. a w-a'l-mah teen-e'
SBJ.2 0-say-PERF me-TOP
le k'iin k'abéet-ech teen-e'
DEM sun/day necessary-ABS.2.SG me-TOP
káa taal-ak-en in w-il-ech
‘you told me that when I would need you I should come to see you’ (HK’AN_452)
The set of fientive verbs is very homogeneous in terms of the derivational operations applicable to
them: to every fientive, a factitive (see §, stative (§5.1.2) and resultative (§5.1.1) derivation
5.2.2 Intransitive verbs from transitive bases
Since there are no verb stems that may be used alternately in transitive and intransitive frames, for
any given root, one of the two valencies is basic, the other one derived.
Transitive roots thus
undergo introversion (§ in order to get their intransitive counterpart, while intransitive roots
undergo extraversion (§ in order to get their transitive counterpart.
Currently, four exceptions are known to this generalization. All of them concern transitive verbs whose introversive
derivation is not recognizable for phonological reasons.
The terms and concepts of introversion and extraversion are introduced in Paris 1985. There are, to the best of my
knowledge, no better terms available (terms like ‘direct object deletion’ being woefully inadequate). However, since
Christian Lehmann, Valency classes in Yucatec Maya 15
There are basically two ways of detransitivizing a transitive verb, by blocking its undergoer (or
direct object) slot and by blocking its actor (or subject) slot. These operations and their variants will
be considered in turn. Introversion
Both basic and derived transitive stems may be introverted. This amounts to transforming the
transitive verb construction of S12, illustrated by E32.a, into the intransitive verb construction of
S9, more specifically, its subtype S19, whose S corresponds to the A of S12 and whose X is
anything other than a direct object. The construction is illustrated by E32.b
S19. Introversion
[ [ Pn
] [ [ W ]
(X) [ S
E32 a. Hwaan-e' t-u ts'ik-ah Peedroh
John-TOP PRFV-SBJ.3 shave-CMPL Peter
‘John shaved Peter’
b. le máak-o' k-u ts'iik
DEM person-D2 IMPF-SBJ3 shave\INTROV
‘that person shaves (people)’
The derivational operator represented by I
in S19 converts its base into an intransitive verb
stem of the active subclass. It has a number of allomorphs which are essentially conditioned by the
basic vs. derived character of W. If W is a primitive transitive root, then the derivational operator is
mostly low tone on the root syllable, as it appears in E32.b. If W is derived, the introversive
operator is a suffix, chiefly -ah, as in E33 (see E72 below for the transitive stem of that verb).
E33 Máax le k-u ka'n-s-ah way-e' ?
‘Who (is the one that) teaches here?’
As anticipated in §4.2.1, the direct object of a transitive verb can always be omitted. Apart from not
mentioning its referent, this has no semantic effect. But introversion is not object omission. The
syntactic effect of the operation is that the verb becomes intransitive, so no direct object can be
combined with it. The semantic effect is that no undergoer is identifiable, which may imply that
there is none.
Though introversion is formally always applicable to a transitive stem, including ditransitive
stems, there are many with which it does not make much sense. For instance, with mach ‘seize,
touch’, thinking up situations where somebody seizes without there being a referent that he seizes is
somewhat artificial. Nevertheless, in the Mayan lexicographic tradition, all transitive verbs are
lemmatized in their introversive form.
Paris 1985 is not written in English, the terms have not made their way into mainstream typology.– Introversion has
been called antipassive in Mayan linguistics. That, however, is a concept proper for an ergative system as it exists in
other Mayan languages, but not in Yucatec.
Christian Lehmann, Valency classes in Yucatec Maya 16 Reflexive construction of transitive verb
All transitive verbs for which it makes sense enter the reflexive construction. It may be derived
from the basic transitive construction S12 by replacing its direct object P by the possessive
construction shown in S20 (see § for indirect reflexivity). E34 is an example.
S20. Reflexive construction
[ [ Pn
[ W ]
[ P
báah ]
[ A
E34 a. Hwaan-e' t-u ts'ik-ah Peedroh
John-TOP PRFV-SBJ.3 shave-CMPL Peter
‘John shaved Peter’
b. le wíinik-o' t-u ts'ik-ah u báah
DEM man-D2 PRFV-SBJ.3 shave-CMPL POSS.3 self
‘the man shaved (himself)’
Báah is an inalienable noun meaning ‘self’. Just as in English myself etc., it combines with the
possessive clitics (P
), cross-referencing the subject A and coding its person and number. The
binary combination forms a possessive NP (S4 without the Pr NP) that constitutes the direct object
of the transitive verb. There is no absolutive suffix on W (cross-referencing j) because that object is
grammatically third person singular even if A is plural.
Although there are shortened forms of this construction, it still exhibits a rather low degree of
grammaticalization. S20 therefore instantiates the transitive schema S12 rather than the intransitive
schema S9. Likewise, it is always fully compositional; there are no formally reflexive verbs devoid
of a coreference relation like German sich schämen ‘feel ashamed’. On the other hand, there is no
simpler way of expressing the meaning of E34.b, no middle voice or the like comparable to English
the man shaved. In particular, the introversive does not code this meaning. Reciprocal construction of transitive verb
All transitive verbs for which it makes sense enter a reciprocal construction. Like the reflexive
construction S20, it is a subtype of the transitive construction S12. S21 formalizes its full form as
illustrated by E35; see § for indirect reciprocity.
S21. Reciprocal construction
[ [ Pn
[ (paaklan) W ]
( [ P
báah ]
) [ A
E35 le máak-o'b-o' k-u paaklan méek'-ik u báah-o'b
DEM person-PL-D2 IMPF-SBJ.3 each.other hug-INCMPL POSS.3 self-3.PL
‘those men hug each other’
Paaklan is an adverb meaning ‘together, each other’. In the former meaning, it may also accompany
intransitive verbs, of course, with no reciprocal effect. It normally takes the position immediately
preceding W indicated in S21, but is susceptible of focusing, so that it precedes Pn, as in E61 and
E62 below. On the other hand, like other adverbs in preverbal position, paaklan optionally forms a
compound with the verb. This may be diagnosed with certainty under the condition that the verb’s
stem is a basic transitive one (s. §4.2.1). As such, it belongs to the conjugation class which lacks a
stem-forming suffix, while as a compound verb, it conjugates via the -t suffix (cf. §
Christian Lehmann, Valency classes in Yucatec Maya 17
The other component of S21 that signals reciprocity is the reflexive phrase already seen in S20.
Just as it does there, it takes the place of W's direct object. Now either of these two markers is
optional in the reciprocal construction. E36 designates a reciprocal situation merely by means of the
reciprocal adverb. In this case, the reciprocal construction is merely an interpretation of a
E36 k-u paaklan púus-t-ik y-ich-o'b
IMPF-SBJ.3 each.other remove.dust-TRR-INCMPL POSS.3-eye-PL
‘they clean each other's face’
And E37 shows a reciprocal situation (or rather, two of them) signaled exclusively by the reflexive
E37 t-u méek'-ah u báah-o'b
PRFV-SBJ.3 hug-CMPL POSS.3 self-PL
úuchik u y-il-ik u báah-o'b
OBL.SR SBJ.3 0-see-INCMPL POSS.3 self-PL
‘they hugged each other when they met’
This is, thus, formally a reflexive construction which is interpreted as reciprocal, as in many other
languages including Spanish. Passive construction
Coming now to detransitivization operations which demote the underlying subject, we start with the
passive. Yucatec has a passive of the common or garden variety, schematized by S22, where P is
the underlying object and A the underlying subject. It is illustrated by E38, which is the passive of
S22. Passive construction
[ [ Pn
] [ [ W ]
[ P
[ tuméen [ A ]
E38 h méek'-ab le chaan xibpal tuméen u maamah-o'
PRFV hug-CMPL.PASS DEM little boy by POSS.3 mother-D2
‘the little boy was hugged by his mother’
The passivized verb belongs to the inactive subclass of intransitive verbs. The operation is fully
productive; W may be any basic or derived transitive stem; and passive constructions are, in fact,
quite frequent in the texts. The passive operator has a few allomorphs. It is chiefly a glottal stop
which is infixed into the host. If the verb stem is a primitive transitive root (s. §4.2.1), the infix goes
into it, as in E33. Otherwise, it goes into the status suffix, as in E57. This is, thus, the only infix on
record which may be inserted in an affix. The agent phrase is optional and occurs with some
frequency in the texts.
Christian Lehmann, Valency classes in Yucatec Maya 18 Deagentive construction
The other operation which affects the actor of the transitive verb is deagentivization
. It applies an
operator (D
) to S12 which blocks its A position and converts its P into subject. The underlying
subject A cannot be accommodated in a deagentive construction. The construction is represented by
S23, a subtype of S9, and illustrated by E39.b, based on a transitive version similar to #a.
S23. Deagentive construction
[ [ Pn
] [ [ W ]
[ P
E39 a. le xch'úupal-o' tu hat-ah hun-waal le analte'-o'
DEM girl-D2 PRFV-SBJ.3 tear-CMPL one-CL.flat DEM book-D2
‘the girl tore a page from the book’
b. h háat in nook'
PRFV tear\DEAG POSS.1.SG cloth
‘my clothes got torn’
W must be a primitive transitive root. The deagentive operator in S23 (D
) materializes as high
tone on the root. The root phonotactics of the language are such that primitive transitive roots are
toneless, but phonologically compatible with a high tone. Non-primitive transitive roots allow no
deagentivization. The derived stem joins the inactive subclass of intransitive verbs. The meaning of
this derivation is that the process in question happens and there is no agent to which reference could
be made; thus, it may happen by itself.
However, there is no emphasis on this; that is reserved for
the spontaneous derivation to be discussed below.
The deagentive also applies to some transitive verbs that involve an instrument. The latter is left
intact by the transformation, as may be seen by comparing E40 with the transitive E20.
E40 k-u píix-il in nal yéetel k'áax
IMPF-SBJ.3 cover\DEAG-INCMPL POSS.1.SG corncob with jungle
lit. ‘my corn plants cover with brush’
The transitive verb stems that do not undergo deagentivization fall into two main formal classes,
non-primitive roots and complex stems. Both of these contain items and subclasses for which a
deagentive derivation would not make much sense, simply because the situations in question are
inconceivable without an actor. The former class includes verbs such as méek’ ‘hug’. The latter
includes, importantly, extraversive bases. For instance, tukult ‘think about’ is the extraversive of
intransitive tuukul ‘think’. The deagentive of the latter would have to mean ‘get thought (of) by
oneself / [for some proposition] to think itself’. Since extraversion presupposes an action and
signals that it extends to an undergoer, deagentivization would, in fact, make little sense with most
extraversive verbs.
On the other hand, there is a class of derived transitive stems for which no deagentive is
morphologically possible although it would make sense. These are transitive stems involving
factitive or causative derivation. However, the constraint does no harm since for these, the bases
themselves fulfill the function of the deagentive form.
The deagentive is called middle in Bricker et al. 1998:346 and Martínez Corripio & Maldonado 2010. It is widely
called anticausative in linguistic typology.
See Martínez Corripio & Maldonado 2010 for some semantic analysis.
Christian Lehmann, Valency classes in Yucatec Maya 19
All transitive verb stems, no matter whether they allow the deagentive derivation, like primitive
transitive roots, or not, like other transitive roots and derived transitive stems, may undergo a
spontaneous derivation (see Lehmann 1993, §6.3.2). This is afforded by a variety of suffixes (see
Bricker et al. 1998:346-349) which are available to transitive stems irrespectively of their form. The
derivation means ‘W happens by itself (spontaneously and unexpectedly)’, as illustrated by E41. It
thus contrasts with the deagentive and emphasizes the spontaneous character of the deagentive
E41 a. t-in tóok-ah in kool
PRFV-SBJ.1.SG burn-CMPL POSS.1.SG milpa
‘I burnt my cornfield’
b. h tóok-k’ah in kool
‘my cornfield burnt (unexpectedly) by itself’
5.3 Formation of transitive verbs
All of the derived constructions introduced in this section are subtypes of the basic transitive verb
construction S12.
5.3.1 Transitive verbs from nominal bases: usative verbs
Besides the major process for the verbalization of nominal bases, viz. the fientive analyzed in
§5.2.1, there is one other such process, which yields transitive verbs from nominal bases. Starting
from the possessive construction S4, one substitutes A for its Pr, and W for its Pd, so it reads ‘A’s
W’. Next W becomes the operand of the usative suffix -int (Bricker 1970). This actually consists of
two operators: -in- verbalizes W, so that A becomes its subject; and -t extraverts it (, so it
takes an undergoer P.
In this way, one obtains the usative construction S24, which may be
paraphrased as ‘A’s W is P = A has/uses P as (his) W’ and, with dynamic aspectuality, ‘A converts
P into (his) W’. E42 is a perfectly compositional illustration of this operation.
S24. Usative construction
[ [ Pn
[ [ W ]
-int ]
[ P
[ A
E42 a. Ba’n a k’aan ?
what POSS.2 hammock
‘what is your hammock?’
b. Ba'n ken a k'áan-int-eh ?
what SR.FUT SBJ.2 hammock-USAT-SUBJ
‘What are you going to use as a hammock (= where are you going to sleep)?’
Apart from exceptional contexts such as E28, the -in- formative does not occur without a following -t; but several
usative verbs, esp. ones based on kinship terms, dispense with the -in-, which is morphologically possible because
extraversion also applies to nominal bases.
Christian Lehmann, Valency classes in Yucatec Maya 20
E43 a. le he’l-a’ u búuk le ko’lel-o’
DEM PRSV-D1 POSS.3 dress DEM adult.woman-D2
‘this is that woman’s dress’
b. le ko'lel-o' t-u búuk-int-ah u nook' u paal
DEM woman-D2 PRFV-SBJ.3 dress-USAT-CMPL POSS.3 clothing POSS.3 child
‘(lit.) that woman used her child’s clothing as a dress = that woman put on her child's
E44 ko'ne'x si'-int-ik hun-kúul che'
go.HORT:AUG firewood-USAT-INCMPL one-CL.plant tree
‘let’s get firewood from a tree’ (EMB_0497)
The derivation applies productively both to alienable (E42.a, E44) and to inalienable (E43.a) bases
(the latter including, importantly, kinship terms). Nevertheless, some usative verbs such as the one
in E43.b are lexicalized.
5.3.2 Transitive verbs from verbal bases
The derivation of transitive verbs from verbal bases may be subdivided according to the valency of
the base. However, derivation of transitive verbs from transitive bases will not be considered as it
does not affect valency. Moreover, there is no productive grammatical process to obtain a transitive
construction by valency reduction of a ditransitive base. A few cases that come close to this
construct will be discussed in §5.3.3. Thus, the bulk of this section concerns the derivation of
transitive stems from intransitive bases. The subdivision (§§ is according to the
actant whose slot is added: first operations adding an undergoer (direct object) slot, then operations
adding an actor (transitive subject) slot are treated. Extraversion
Given a base of the active intransitive subclass, an extraversive stem may be formed by appending
the suffix -t (glossed T
‘transitivizer’) to it. The operation transforms the intransitive construction
of S9, illustrated by E12, into the extraversive construction of S25, illustrated by E45. S25 is a
subtype of the basic transitive construction S12 and a mirror-image to introversion (S19).
S25. Extraversion of a verbal base
[ [ Pn
[ [ W ]
-t ]
[ P
[ A
E45 le x-ch’úuppal-o' k-u síit'-t-ik le koot-o'
DEM F-girl-D2 IMPF-SBJ.3 jump-TRR-INCMPL DEM wall-D2
‘the girl jumps (over) that wall’
As discussed more fully in §, the same derivation, formally, applies to bases of other word
classes, too. Verbal bases, however, obey without exception the constraint that they must be active.
This includes compound verb stems, especially incorporative verbs, all of which belong to the
active conjugation class irrespectively of the class of their verbal core. For instance, the stem háak
‘slide’ is inactive. The incorporative stem háakchek’ ‘slide stepping on something’ illustrated in
E46.a is active and thus allows the extraversion of #b (see §6.2 on retransitivization of incorporative
Christian Lehmann, Valency classes in Yucatec Maya 21
E46 a. háak-chek’-nah-en
‘I slipped (by stepping on something)’
b. t-in háak-chek’-t-ah le ha’s-o’
PRFV-SBJ.1.SG slide-by.foot-TRR-CMPL DEM banana-D2
‘I slipped on that banana’
Extraversion may be conceived as a lexicalized variant of the applicative operation. It differs in a
variety of respects from the canonical applicative (Lehmann & Verhoeven 2006):
It is a lexical rather than syntactic operation, which entails that it is neither compositional nor
fully productive.
It only applies to intransitive bases, while an applicative may freely apply to transitive bases.
Consequently, extraversion is always valency-increasing, while the applicative is sometimes
just valency-rearranging.
The applicative promotes a clause component low on the hierarchy of syntactic functions to
direct object function. Extraversion does not promote anything, since the resulting direct object
typically cannot even be accommodated in the underlying intransitive construction. The specific
role of the undergoer exteriorized from the verb depends essentially on the latter’s lexical
Thus, in the intransitive construction of E47.a, there is no way of accommodating an addressee,
which is, however, the role of the direct object in the extraversive construction of #b.
E47 a. le xibpal-o’ túun xóob (*ti’ le x-ch’úuppal-o’)
DEM boy-D2 PROG\SBJ.3 whistle LOC DEM F-girl-D2
‘the boy is whistling (at the girl)’
b. le xibpal-o’ túun xóob-t-ik le x-ch’úuppal-o’
DEM boy-D2 PROG:SBJ.3 whistle-TRR-INCMPL DEM F-girl-D2
‘the boy is whistling at the girl’
On the other hand, extraversion is formally completely regular and highly productive for active
intransitive bases, elementary or complex. For these, there is an alternative to its conception as a
derivational process. Instead, one might abide by the two conjugation classes of transitive verbs
explained in T1, the first without a stem-forming exponent, only for basic transitive roots, and
another with the stem-forming suffix -t, both for derived transitive stems (such as háakchek’ ‘slide
stepping on something’ in E46) and such roots as have an intransitive active use (such as xóob
‘whistle’ in E47). Then there would be no process of extraversion, and instead verbs like xóob and
háakchek’ would be ambitransitive, i.e. alternating between transitive and active-intransitive
valency. While this might be sufficient for verbal bases, there remains the verbalization of nominal
bases such as cha’n ‘spectacle [n.]’ cha’nt ‘watch’, which, under this alternative account, would
have to be conceived as a conversion (a zero-operator recategorization) with transfer into the second
conjugation class. However, conversions normally only change the category without affecting the
meaning, while extraversion may yield unforeseeable semantic results, as in E47. Another argument
in favor of the extraversion analysis is the alternation between the -t and the -s suffix (s. §,
the latter of which is an unequivocal derivational operator.
On the other hand, the same morphological process of extraversion also applies to transitive
verboids in order to verbalize them. To formalize this, it suffices to replace the category subscript of
W in S25 by ‘transitive verboid’. The process is illustrated by E48.
Christian Lehmann, Valency classes in Yucatec Maya 22
E48 a. le xibpal-o' u k'ahóol le xch'úupal-o'
DEM boy-D2 SBJ.3 know DEM girl-D2
‘the boy knows/knew the girl’
b. le xibpal-o' k-u k'ahóol-t-ik le xch'úupal-o'
DEM boy-D2 IMPF-SBJ3 know-TRR-CMPL DEM girl-D2
‘the boy knows/knew the girl’
Just as in the case of the fientive verb (§5.2.1), the primary purpose of this operation is to make the
concept available in the verbal category. In the present case, the base already has the same valency
as the derived verb. The verbalization of transitive verboids is, therefore, another argument in favor
of the alternative analysis of the -t suffix as a conjugation class suffix instead of an operator of
extraversive derivation. The indeterminacy of this analysis is due to the fact that languages may use
the same formative both as a derivational operator and as an inflection-class exponent. Factitive verbs
To every deadjectival fientive verb (§5.2.1) there corresponds a factitive verb. It is constructed as
schematized in S26, a subtype of the basic transitive construction S12, and illustrated in E49.
S26. Factitive derivation
[ [ Pn
[ [ W ]
-ki/unt ]
[ P
[ A
E49 t-in tikin-kunt-ah le sikil-a'
PRFV-SBJ.1.SG dry-FACT-CMPL DEM pumpkin.seed-D1
‘I dried this pumpkin seed’
E49 is based on a sentence like E6 and is the factitive counterpart to the fientive E30.b. The
factitive is that variant of the causative whose base is adjectival. The suffix is morphologically
complex, as it consists of the factitive suffix proper, -kin/kun, plus the extraversive suffix -t. The
first component -kin/-kun does not occur alone in intransitive use, but it does occur in the citation
form of these verbs and in their nominalization. Depending on lectal variation, the causative suffix
-s appears instead of the -t, yielding the variant -kins/kuns, as in E51.
Just like any adjective, the adjectival experiential construction of S6 may be factitivized. Thus,
E50.b is a factitive construction on the basis of #a.
E50 a. ki’mak in w-óol
happy POSS.1.SG 0-mind
‘I am happy’
b. le paax-o' k-u ki'mak-kuns-ik in w-óol
DEM music-D2 IMPF-SBJ.3 happy-FACT-INCMPL POSS.1.SG 0-mind
‘that music delights me’
Furthermore, factitives are regularly formed from positional bases. E51 is the factitive counterpart
to fientive E14.
E51 Kul-kins y-óok'ol le k'áan-che'-o' !
sit-FACT(IMP) 0-over DEM hammock-wood-D2
‘Make him sit (down) on the stool!’
Christian Lehmann, Valency classes in Yucatec Maya 23
E51 follows the factitive pattern of S26 and the ditransitive pattern of transport verbs formalized as
S15. It is, however, noteworthy that the factitive rather than the causative operator should be used
with positional bases. Positional roots are clearly verbal rather than adjectival, as proved both by
their conjugation and by the necessity of applying derivational operators such as the stative and
resultative if one needs them in the adjectival category (§5.1). Positional concepts are notoriously
ambivalent as to the alternative of ‘stative (and uncontrolled)’ vs. ‘dynamic (and controlled)’. If we
assume a scale of dynamicity as set out in Lehmann 1993, §3.1, then positionals may be allocated a
point between state and process. The factitive derivation operates on bases which designate
properties and states (adjectival bases) and bases which designate positions, and would thus be
operating on a set of bases which together cover a continuous segment of the dynamicity scale. Causative of inactive intransitive verb
Given a base of the inactive intransitive subclass, a causative stem may be formed. The operation
transforms the intransitive construction of S9, illustrated by E13, into the causative construction of
S27, illustrated by E52. S27 is a subtype of the basic transitive construction S12.
S27. Causative construction
[ [ Pn
[ [ W ]
[ P
[ A
E52 t-in lúub-s-ah le nuxib che'-o'
PRFV-SBJ.1 fall-CAUS-CMPL DEM old wood-D2
‘I felled that old tree’
A is the causer, P is the causee. The causative suffix has a set of allomorphs, the basic one of which
is the -s appearing in E52. With very few exceptions, W must be inactive. Consequently,
intransitive verbs of the active subclass, e.g. meyah ‘work’, cannot be causativized. Fientive stems
cannot be causativized, either; but that is not necessary, since the corresponding function is fulfilled
by applying the factitive derivation to their base. Finally, transitive stems cannot be causativized.
This constraint may be circumvented by first passivizing them, thus creating an inactive derived
stem. However, only one causative verb, ka’ns ‘teach’, discussed in §6.1, is known to be formed by
such a procedure. There are periphrastic constructions which apply to such bases as are not
amenable to the causative derivation.
Two intransitive constructions should be singled out as undergoing causativization. The first is
the experiential construction sketched in §4.1.2. In causativizing S10, óol ‘mind’ with its possessor
becomes the direct object, and L of S10 disappears. E53 illustrates the causativization of the verb
appearing in E16.
E53 he'l a háak'-s-ik in w-óol-e'
DEF.FUT SBJ.2 scared-CAUS-INCMPL POSS.1.SG 0-mind-D3
‘you will scare me’ (kuruch_021)
The other intransitive construction whose causativization is worth mentioning is the directed motion
verb construction briefly described in §4.1.3. In causativizing it, the verb of motion becomes a verb
of transport and S of S11 becomes P of S27. E54 is a rather close causative counterpart to
intransitive E17.
Christian Lehmann, Valency classes in Yucatec Maya 24
E54 Hóok'-es le peek' ich nah-o'!
leave-CAUS DEM dog in house-D2
‘Get (chase) the dog out of the house!’
Remember that verbs of directed motion are inactive; their causativization is thus completely
regular. And in fact, the entire set can be causativized. The causatives of the two most basic ones,
bin ‘go’ and taal ‘come’, are bis ‘transport, take to’ and taas ‘bring’. Their lexicalization goes hand
in hand with phonological erosion (from *bin-s and *taal-s). In general, the verb of transport
inherits the local complement from the underlying verb of motion.
Two exceptions are known to the rule that causatives are only formed from inactive intransitive
bases: active intransitive áalkab ‘run’ and balak’ ‘roll’ form the causatives áalkabans and balak’es.
These are, at the same time, the only verbs that allow both causative and extraversive
transitivization: áalkabt ‘run with respect to’, balak’t ‘roll [sth.]’. Since this coexistence is not
provided for by the system, it leads to problems of usage in both cases: speakers constantly confuse
áalkabt and áalkabans (cf. §; and balak’es and balak’t are synonymous, while in general the
suffix -t does not have causative value.
Given the constraints on factitive and on causative derivation, this pair of derivations together
covers a continuous segment, viz. the lowest segment, of the implicational hierarchy of
causativization proposed in Lehmann 2013. Functional ambiguity of transitivizer
In terms of verbal morphology, there are three kinds of transitive verb stems:
1) transitive roots, e.g. kach ‘break’
2) causativized verb stems, derived by one of the allomorphs of the causative suffix -s, e.g.
na’k-s ‘make get up, lift’
3) other derived verb stems, derived by the suffix -t, e.g. ts’alk’ab-t ‘stamp [sth.] with the
Since both #1 and #2 are closed classes, the vast majority of transitive verbs in the language
bear the suffix -t. Its functions may be classified as follows:
a) It derives extraversives from active intransitive roots, as in áakan ‘moan’ áakant
‘bemoan’, che’h ‘laugh’ – che’ht ‘laugh at’, ts’iib ‘write’ ts’iibt ‘write [sth.]’, etc. See
b) It derives transitive verbs from non-verbal roots, as in muuk’ ‘force [n.]’ muuk’t ‘force
[sth.]’, cha’n ‘spectacle’ – cha’nt ‘look at’.
c) It “retransitivizes” incorporative verbs, as in ts’al-k’ab ‘stamp with hand’ – ts’alk’abt
‘stamp [sth.] with one’s hand’. See §6.2.
d) Finally, -t is a morphological component of the usative suffix –int (§5.3.1) and the factitive
suffix -kint/-kunt (§
This plurifunctionality of the -t suffix leads to the consequence that it sometimes takes the place
of the causative suffix. The intransitive use of áalkab ‘run’ is illustrated by E55.a. E55.b illustrates
a regular extraversive, #c a regular causative derivation.
E55 a. le tsíimin-o' áalkab-nah ti' le beh-o'
DEM horse-D2 run-CMPL LOC DEM path-D2
‘the horse ran along the path’
Christian Lehmann, Valency classes in Yucatec Maya 25
b. le tsíimin-o' t-u y-áalkab-t-ah le beh-o’
DEM horse-D2 PRFV-SBJ.3 0-run-TRR-CMPL DEM way
‘that horse ran the path’
c. t-in w-áalkab-ans-ah le tsíimin-o'
‘I raced the horse’
In certain contexts, the extraversive and the causative derivation of áalkab are semantically not too
different. Examples like E56 might therefore provide the context for reinterpretation of the
extraversive suffix. Here the causative and the extraversive suffix alternate in causative function.
E56 le peek'-o'b-o' k-u y-áalkab-t-ik-o'b le k'éek'en-o'b-o'
DEM dog-PL-D2 IMPF-SBJ.3 0-run-TRR-INCMPL-3.PL DEM pig-PL-D2
‘the dogs run behind the pigs / make the pigs run’
To complete this picture, the factitive suffix -kint/-kunt has a variant -kins/-kuns, containing the
causative instead of the extraversive morpheme.
On the basis of the construction of the adjectival experiential predicate illustrated by E8, the
transitive verb stem ok’om-óolt is formed. In E57.a, it is used with an extraversive argument
structure; in #b, it has a causative argument structure.
E57 a. h ok'om-óol-t-a'b úuchik u kíim-il
‘he was mourned when he died’
b. le ba'l-o' k-u ok'om-óol-t-ik-en
DEM thing-D2 IMPF-SBJ.3 sad-mind-TRR-INCMPL-ABS.1.SG
‘that thing makes me sad’
In the latter case, a factitive derivation would be expected, which is, however, nonexistent for this
verb. Thus, the extraversive suffix takes on the function of the causative suffix, developing into a
generic transitivizer.
5.3.3 Transitive verbs from ditransitive bases
Just as intransitive verbs may be derived from transitives by valency reduction, so in principle,
transitive verbs might be derivable from ditransitive verbs by valency reduction. However, the only
available processes in this domain, viz. reflexivization and reciprocization, do not, in Yucatec,
actually reduce the valency of the verb, but rather fill the respective actant position with a reflexive
phrase. They will nevertheless briefly be illustrated here. Indirect reflexivity
In a way analogous to the reflexive construction of a transitive verb (§, the indirect object
position of a ditransitive verb (S15) may be filled by the reflexive phrase P
báah ‘X’s self’.
E58 is such a sentence, where the indirect object position would normally be filled by an NP of
disjoint reference.
E58 k-u ch'a'-ik sahak-il ti' u báah
IMPF-SBJ.3 take-INCMPL afraid-ABSTR LOC POSS.3 self
‘he is afraid of himself’
Christian Lehmann, Valency classes in Yucatec Maya 26 Indirect reciprocity
In a way analogous to the reciprocal construction of a transitive verb (§, the indirect object
position of a ditransitive verb (S15) may be filled by the reflexive phrase P
báah in a reciprocal
construction. E59 is such a sentence, where the indirect object position would normally be filled by
an NP of disjoint reference.
E59 t-u paaklan túuxt-ah lool-o'b ti' u báah-o'b
PRFV-SBJ.3 each.other send-CMPL flower-PL LOC POSS.3 self-PL
‘they sent each other flowers’
E60 le paal-o'b-e' k-u ta'k-ik u báah-o'b
DEM child-PL-D3 IMPF-SBJ.3 hide-INCMPL POSS.3 self-PL
ti' u báahtsil-o'b
‘the children hide from each other’
A variant of this is illustrated by E60. It features báahtsil instead of báah, whose formation and
function remain obscure.
Alternatively and just as in the reciprocal construction of transitive verbs, the reflexive phrase
may be omitted. For instance, E61 may be formed on the basis of a sentence like E23.
E61 ichil to'n-e' paaklan k tsikbal-t-ik úuchben tsikbal-o'b
inside us-TOP each.other SBJ.1.PL talk-TRR-INCMPL old story-PL
‘in our group we tell each other old stories’
This then does appear to reduce the valency of the ditransitive verb by suppressing the indirect
object. However, the indirect object is optional, anyway, and not marked on the verb, either.
Consequently, E61 is just a variant of the ditransitive verb construction S15.
Finally, with the communication verbs a’l ‘say’ and k’áat ‘ask’, the preposition of the indirect
object may be omitted under reciprocity, so that it appears as a direct object, as in E62.
E62 in taatah-o'b-e' paaklan u k'áat-ik u báah-o'b taak'in
POSS.1.SG father-PL-TOP each.other SBJ.3 ask-INCMPL POSS.3 self-PL money
‘my parents ask each other for money
These data show, if anything, that reciprocity of non-direct objects is not well-established in the
6 Other alternations
This section brings together a set of valency-related alternations which are not coded on the verb.
The actant shift to be discussed in §6.1 is hardly coded at all. The other two subsections deal with
two argument structure alternations, viz. indirect participation and incorporation, the first of which
is nominally coded, while the latter appears to escape such a classification. Quite generally, it may
be anticipated here that there are very few uncoded alternations in Yucatec, and they are
The derivation in -tsil from a relational base normally forms a non-relational counterpart to it (cf. Lehmann 2002, ch. That is, however, obviously not what happens in the case of báahtsil.
Christian Lehmann, Valency classes in Yucatec Maya 27
unsystematic. This fits in with the earlier observation that in this language, valency is a property
which is firmly associated with a verb stem and whose change requires a morphological operation.
6.1 Shifts in three-argument verbs
There is a set of verbs whose meaning involves three arguments of which normally no more than
two surface in the form of actants. The question then arises how the non-first arguments are mapped
onto non-subject functions. There is no general strategy for this problem, but two partial ones. The
first of them, called indirect participation, maps the second argument onto the direct object and
accommodates the third argument not as an actant of the verb, but as a dependent of the second
actant. Under semantic conditions to be specified in §6.1.1, it works in a relatively regular way. The
other strategy, called direct object shift, maps either of the non-first arguments on the direct object
function, while the other is preferably omitted or, at most, adjoined in a prepositional phrase. This
strategy is applicable to a small set of verbs to be reviewed in §6.1.2 and takes rather idiosyncratic
shapes for each of them.
6.1.1 Indirect participation
The nominal possessive construction S4 is much more central to Yucatec syntax than the trivalent
verb construction S14. Given a three-participant situation with a lexically coded (rather than
pronominal or implicit) undergoer, if a participant bears a relation to this undergoer which can be
construed as possessive in some sense, then a construction in which that participant functions as
possessive attribute of the undergoer is often preferred to a construction in which it is the third
verbal actant. That amounts to using S28 as a variant of S14 (with identity of all variables).
S28. Indirect participation
[ [ Pn
[ W ]
[ P
P [ L
[ A
This construction is called indirect participation because although L is a participant in the situation
whose core is coded by W, L is not coded as a dependent of W, and instead its relation to W is
mediated by its possessee P.
In E63 and E64, the #a versions are constructed trivalent versions of
the #b versions, which are attested in the corpus. The brackets enclose an NP where the #a versions
show an NP and a PrepP each directly depending on the verb.
E63 a. máantats' táan u t'ab-ik kib ti’ kili'ch Anton
constantly PROG SBJ.3 lighten-INCMPL candle LOC saint Anton
b. máantats' táan u t'ab-ik u kib kili'ch Anton
constantly PROG SBJ.3 lighten-INCMPL [POSS.3 candle saint Anton]
‘regularly he lightens candles for St. Anthony’
This is also true of the experiential construction schematized by S8 above. See the account in Lehmann et al. 2000
and Malchukov et al. 2010, §3.2.3.
Christian Lehmann, Valency classes in Yucatec Maya 28
E64 a. káa t-u máan-s-ah u éerensyah t-u ìihoh-e'
heritage LOC-POSS.3
b. káa t-u máan-s-ah u éerensyah u ìihoh-e'
heritage POSS.3
‘and he
handed his
heritage to his
Indirect participation, as in the #b versions, is the traditional construction both for benefactive and
similar adjuncts, as in E63.b, and for third arguments, as in E64. Increasing contact with Spanish
has meanwhile established the #a versions as alternatives. This is, thus, a valency alternation in
which a verbal dependent alternates with a nominal dependent.
Another two-place verb which accepts a beneficiary in indirect participation similarly as t’ab in
E63 is huch' ‘grind’. Other three-argument verbs which show the same alternation as máans in E64
include ts’a’ ‘put, give’, taas ‘bring’, okol ‘steal’ and but’ ‘stuff’. It deserves particular attention
that even the central three-place verb of the language, viz. ts’a’, is frequently found in the indirect
participation construction. E65 is a representative example from a tale.
E65 yuum ahaw-e'
master/father chief-TOP
káa bin t-u ts'a'-ah u y-otoch x-t'uup
CONJ QUOT PRFV-SBJ.3 put/give-CMPL [POSS.3 0-home F-youngest.sibling]
‘the chief gave a house to the youngest daughter’ (HK'AN_309)
Moreover, there is a set of three-argument verbs for which no trivalent construction is possible,
indirect participation being the only construction allowing the simultaneous coding of all three
arguments. This set includes lak ‘detach’, luk’s ‘take away (from)’ (causative of luk ‘depart, go
off’), hat ‘tear (off)’ (illustrated by E39.a), tix ‘rinse (off)’, tíit ‘shake (off), púus-t ‘remove dust’,
ts'ik ‘shave’ and ts’íil ‘peel’. As may be seen, the set is semantically homogeneous, as they all mean
‘A affects an object L in the role of source in such a way that another object P in the role of moved
undergoer gets detached from L’ (variables as in S28). In the indirect participation construction, L
appears as the possessive attribute of P, as shown in S28. For the verbs lak and luk’s, P is the only
possible direct object. The other verbs in the set show direct object shift (s. next section) with
respect to P and L, i.e. they allow an alternative construction in which L is the direct object, while P
cannot be accommodated. E66 and E67 illustrate this alternation for ts’íil ‘peel’ and púus-t ‘remove
dust’: the #a version has indirect participation with P in direct object function, while the #b version
has L in direct object function.
E66 a. le xibpal-o' t-u ts'íil-ah u sóol le che'-o'
DEM boy-D2 PRFV-SBJ.3 peel-CMPL [POSS.3 shell DEM tree-D2]
‘the boy peeled the bark off the tree
b. ts'íil-a'b tuláakal le che'-o'b-o'
peel-PASS.CMPL all DEM wood-PL-D2
‘all the trees were peeled’
E67 a. le xch'úup-o'b-o' t-u púust-ah-o'b u luuk'-il le meesah-o'
DEM woman-PL-D2 PRFV-SBJ.3 wipe-CMPL-PL [POSS.3 mud-REL DEM table-D2]
‘the women wiped the dirt off the table’
Ts’ik is somewhat exceptional in that L is an animate being, as one can shave either a person or his hair or beard.
Christian Lehmann, Valency classes in Yucatec Maya 29
b. le xch'úup-o'b-o' t-u púust-ah-o'b le meesah-o'
DEM woman-PL-D2 PRFV-SBJ.3 wipe-CMPL-PL DEM table-D2
‘the women wiped (dirt off) the table’
While the independent existence in the universe of discourse of a possessive semantic relation
between P and L favors indirect participation, there are many examples where a possessive
syntactic relation owes its entire raison d’être to this construction. For instance, in E67.a, the table
in no sense possesses the dirt. Instead, the syntactic possessor function exclusively serves the
coding of a verb argument role.
6.1.2 Direct object shift
With a set of Yucatec three-argument verbs, either of the non-first arguments may be direct object,
while the other non-first argument is preferably omitted. This alternation is called direct object shift.
The set may be further subdivided as follows (with variables as in S14):
1. only bivalent (monotransitive) construction possible: hat ‘tear P off L’, tix ‘rinse P off L’, tíit
‘shake P off L’, púus-t ‘remove P (dust) from L’, ts'ik ‘shave L’s P’ and ts’íil ‘peel P off L’;
2. trivalent construction possible:
a. with locative shift: but’ ‘stuff P into L, fill L with P’, bak’ ‘wind P around L, wrap L in P’,
b. with irregular valency alternation: okol ‘steal P from L’, ka’ns ‘teach L P’.
The subset #1 is the same mentioned in the preceding section as the set of verbs allowing
indirect participation of L if P is direct object; s. E66f. With verbs of subset #2, the third argument
is introduced by a preposition. With subset #2a, S14 (V P ti’ L) alternates with S13 (V L yéetel P).
That is: if P is the direct object, as in E68.a and E69.a, then L is coded in a local prepositional
phrase. If, however, L is the direct object, as in the #b versions, then P appears as an instrumental
adjunct. This pattern defines locative shift.
E68 a. le koolnáal-o' t-u but'-ah ixi'm ti' le kaamion-o'
DEM farmer-D2 PRFV-SBJ.3 fill-CMPL corn LOC DEM truck-D2
‘the farmer loaded corn onto the truck’
b. le xch'úupal-o' t-u but'-ah le luuch yéetel ha'-o'
DEM girl-D2 PRFV-SBJ.3 fill-CMPL DEM cup with water-D2
‘the girl filled the cup with water’
E69 a. k-in bak'-ik su'm t-in k'ab
IMPF-SBJ.1.SG wind-INCMPL rope LOC-POSS.1.SG hand
‘I wind a rope around my hand/arm’
b. k-in bak'-ik in k'ab yéetel su'm
IMPF-SBJ.1.SG wind-INCMPL POSS.1.SG hand with rope
‘I wrap my hand/arm with a rope’ (FEE_0131)
These two are, then, the only Yucatec verbs on record displaying locative shift. The #b examples
were elicited; these constructions do not occur in the corpus.
Subset #2b comprises an idiosyncratic remainder. The verb okol ‘steal’ may be used in
intransitive, monotransitive and ditransitive constructions. Although there are, in fact, two variant
transitive stems of this base (with or without the -t suffix), neither of them is firmly associated with
either the stolen thing or the deprived person as the direct object. E70 illustrates the variation, with
#a having the stolen thing, and #b, the bereft person, in direct object function.
Christian Lehmann, Valency classes in Yucatec Maya 30
E70 a. t-in w-okl-ah hun-p'éel bisikleetah
PRFV-SBJ.1.SG 0-steal-CMPL one-CL.INAN bicycle
‘I stole a bike’
b. ko'x okl-ik le máak-o'
go.HORT steal-INCMPL DEM person-D2
‘let’s rob that person’
In the ditransitive construction illustrated by E71, the stolen thing is the direct, the victim the
indirect object. This construction – an instantiation of S15 – exhibits no syntactic variation.
E71 k-in w-okol-ik teech le ba'l-o'
IMPF-SBJ.1.SG 0-steal-INCMPL you DEM thing-D2
‘I steal that thing from you’
The most complicated verb in this respect is ka’ns ‘teach’. Its stem is ultimately based on the
transitive root kan ‘learn’. This is first passivized, yielding ka’n ‘be learnt’. This stem is then
causativized, yielding ka’ns ‘cause to be learnt’. By its formation, this verb should take the subject
matter learnt as its direct object. This does happen in E72.a, where the learner appears as indirect
object. In #b, these two participants swap their syntactic functions, something that happens with no
other ditransitive verb. And in #c, the learner is direct object, while the subject matter is apparently
secondary object, a function otherwise unknown in Yucatec.
E72 a. t-in ka'ns-ah xokp'éelil-o'b t-in paal
PRFV-SBJ.1.SG teach-CMPL number-PL LOC-POSS.1.SG child
‘I taught my child numbers’
b. le ko'lel-o' t-u ka'ns-ah le xch'úupal
DEM woman-D2 PRFV-SBJ3 teach-CMPL DEM girl
ti' hum-p'éel k'aay-o'
LOC one-CL.INAN song-D2
‘the lady taught the girl a song
c. ka'ns-a'b-en utsil t'aan maaya tumen in kaanbesah
teach-PASS.CMPL-ABS.1.SG well speak maya by POSS.1.SG teacher
‘I was taught to speak Maya well by my teacher’
Finally, this verb may also be used monotransitively, with either the subject matter or the learner as
direct object. All of this allows of no generalization and instead points to the peculiar idiosyncratic
nature of this verb.
6.2 Incorporation
The incorporative construction must be seen in the context of the formation of complex verb stems.
There are essentially two processes of forming compound verb stems, by combining a verb stem
with a preverbal adverb and by combining it with a postverbal noun. The latter is traditionally
called incorporation and may be schematized as in S29.
S29. Incorporative construction
[ [ Pn
[ [ W ]
– [ P ]
[ A
Christian Lehmann, Valency classes in Yucatec Maya 31
Although P is not necessarily the underlying direct object, as we shall see in a moment, W must be
transitive. Moreover, it must be a transitive root. The only morphological complexity allowed for W
is reduplication (as in ch’a’-ch’a’-book RED-take-odor ‘sniff’) and the distributive suffix -lan (as in
chuk-lan-paach catch-D
-back ‘hunt down separately’). Many incorporative constructions bear
close paradigmatic correspondence with a syntactic construction that has P as a dependent NP or
PrepP. To that extent, it seems justified to treat incorporation not only as a process of verbal
compounding, but also as a syntactic process.
The syntactic function of P in the corresponding verbal dependency construction is either direct
object or instrumental adjunct (see Lehmann 2006 for details). In the former case, S29 corresponds
paradigmatically to S12 (with identity of all variables). E73 shows P first in direct object function,
then incorporated.
E73 a. t-in ch’ak-ah xa’n behela’-ak-e’
PRFV-SBJ.1.SG cut-CMPL palm today-past-D3
‘I cut palm fronds today’
b. h bin-en ch'ak-xa'n behela'-ak-e'
PRFV go(CMPL)-ABS.1.SG cut-palm today-past-D3
‘I went palmcutting today
If the incorporated noun has the semantic role of an instrument, the incorporative construction
corresponds to S13 (with I in S13 mapping onto P in S29). This is illustrated by E74.
E74 a. táan u páan-ik hun-p’éel ba’l yéetel u k’ab le tsíimin-o’
PROG SBJ.3 dig-INCMPL one-CL.INAN thing with POSS.3 hand DEM horse-D2
‘that horse is digging something out with his hoof’
b. táan u páan-k’ab le tsíimin-o’
PROG SBJ.3 dig-hand DEM horse-D2
‘that horse is pawing at the ground’ (Bricker et al. 1998 s.v. páan)
c. táan u páan-k’ab-t-ik hun-p’éel ba’l le tsíimin-o’
PROG SBJ.3 dig-hand-TRR-INCMPL one-CL.INAN thing DEM horse-D2
‘that horse is digging something out (with his hoof)’
No matter what the role of the incorporated noun is, the incorporative verb joins the active subclass
of intransitive verbs, illustrated by E73.b and E74.b. However, many of these verbs are not
generally used intransitively and instead are extraverted in order to take a direct object. That is
shown by E74.c (cf. also E46 above). If the incorporated noun is an instrument, then the direct
object of the incorporative verb may be the same as with the simple verb, as is the case in E74. To
that extent extraversion of the incorporative stem amounts to its retransitivization (Sullivan 1984).
E75 is another example of this.
E75 a. k-u lom-ik yéetel u k'ab
IMPF-SBJ.3 stab-INCMPL with POSS.3 hand
‘he pricks it with his finger’
b. k-u lom-k'ab-t-ik
IMPF-SBJ.3 prick-hand-TRR-INCMPL
‘he pricks it with his finger’
Christian Lehmann, Valency classes in Yucatec Maya 32
If, however, the incorporated noun functions as the undergoer of the base verb, then
retransitivization of the incorporative verb allows for its combination with a new kind of direct
object. In E76, this is the same participant that was a prepositional adjunct in the base version.
E76 a. t-u t'in-ah u y-ich ti' teen
PRFV-SBJ.3 extend-CMPL POSS.3 0-eye LOC me
‘he greeted me opening his eyes’
b. t-u t'in-ich-t-ah-en
PFRV-SBJ.3 extend-eye-TRR-CMPL-ABS.1.SG
‘he winked at me’
On the one hand, not all incorporative verbs have a natural free syntactic counterpart; the #a
versions of E74 – E76 are not very idiomatic. On the other hand, incorporative constructions are not
fully productive in the syntactic sense. Only such instruments and such undergoers are incorporated
which typically figure in the action described by the verb. Thus, for a verb to “regularly
incorporate its instrument does not mean that any instrumental adjunct may be incorporated, but
rather that the incorporative constructions of the verb are morphologically and semantically regular
7 Conclusion
Contrary to nominal valency, which is quite intricate in Yucatec, verbal valency is typologically
unremarkable in many respects: The verb finds its place in a part-of-speech system which is not
essentially different from an SAE system, including the conversion operations between the
categories. The alignment of syntactic functions is accusative without any split; it is only the
morphology of the cross-reference markers on intransitive verbs which displays an ergativity split
depending on certain conjugation categories. There is an indirect object, although somewhat
underdeveloped in comparison with SAE languages. There are valency alternations between
transitive and intransitive frames, including a completely regular passive. The typologically
noteworthy features are the following:
1. The transitivity system is extremely rigid, in the following sense: Every verb form that
occurs in a text is formally either transitive or intransitive, and its syntactic construction
coincides with this. There is no way of using a given verb form in the other function.
2. As a corollary to observation #1, the language has no regular or productive uncoded valency
3. Conjugation classes reflect relationality and control rather faithfully and are correspondingly
productive as targets of derivational operations.
4. The language has an aversion against multivalent constructions and, in fact, against
accumulating dependents on a single verb. On the one hand, it has strategies of avoiding
this. One of these is indirect participation, which employs the nominal possessive
construction instead of verbal dependency. Another strategy with similar effect is the
incorporation of non-referential dependents in the verb. On the other hand, there is not a
single productive operation that produces a ditransitive construction.
Christian Lehmann, Valency classes in Yucatec Maya 33
a. In construction formulas
A actor
Adj adjective
Advl adverbial (phrase)
Aux auxiliary
DC deictic clitic
E experiencer
I instrument
L local and other complement
N noun
Nom nominal
NP noun phrase
P undergoer
PC pronominal clitic
Pd possessed
Pn pronominal element
Pr possessor
PrepP prepositional phrase
S intransitive subject
V verb
VC verbal complex
VCC verbal clause core
b. In interlinear glosses
0 [no meaning]
1, 2, 3 1
, 2
, 3
ABS absolutive cross-reference
ABSOL absolutive
ABSTR abstract
AN animate
AUG augmented number
CAUS causative
CL classifier
CMPL completive
COLL collective
CONJ conjunction
D1/2/3 proximal/distal/anaphoric deictic
DEAG deagentive
DEF definite
DEM demonstrative
F feminine
FACT factitive
FIENT fientive
FUT future
GER gerund
HORT hortative
IMP imperative
IMPF imperfective
INAN inanimate
INCMPL incompletive
INTROV introversive
LOC locative
OBL oblique
PASS passive
PL plural
POS position
POSS possessive cross-reference
PRFV perfective
PROG progressive
PRSV presentative
QUOT quotative
REL relational
RSLTV resultative
SBJ subject cross-reference
SG singular
SPONT spontaneous
SR subordinator
SUBJ subjunctive
TERM terminative
TOP topic
TRR transitivizer
USAT usative
Bricker, Victoria R. 1970, “Relationship terms with the usative suffix in Tzotzil and Yucatec Maya”. Papers
from the Regional Meeting of the Chicago Linguistic Society 6: 75-86.
Bricker, Victoria R. et al. 1998, A dictionary of the Maya language as spoken in Hocabá, Yucatán. With a
botanical index by Anne S. Bradburn. Salt Lake City: University of Utah Press.
Christian Lehmann, Valency classes in Yucatec Maya 34
Lehmann, Christian 1992, “Yukatekische lokale Relatoren in typologischer Perspektive”. Zeitschrift für
Phonetik, Sprachwissenschaft und Kommunikationsforschung 45:626-641.
Lehmann, Christian 1993, “Predicate classes in Yucatec Maya”. Función 13/14:195-272.
Lehmann, Christian 2002, Possession in Yucatec Maya. Second, revised edition. Erfurt: Seminar für
Sprachwissenschaft der Universität (ASSidUE, 10).
Lehmann, Christian 2006, “Incorporación nominal, posesión y participación en el maya yucateco”.
Península 1(2):129-141.
Lehmann, Christian [to app.], “Latin causativization in typological perspective”. Lenoble, Muriel & Longrée,
Dominique (eds.), Actes du 13ème Colloque International de Linguistique Latine. Louvain: Peeters.
Lehmann, Christian & Shin, Yong-Min & Verhoeven, Elisabeth 2000, Direkte und indirekte Partizipation.
Zur Typologie der sprachlichen Repräsentation konzeptueller Relationen. München: LINCOM Europa
(LINCOM Studies in Language Typology, 4).
Lehmann, Christian & Verhoeven, Elisabeth 2006, “Extraversive transitivization in Yucatec Maya and the
nature of the applicative. Kulikov, Leonid & Malchukov, Andrej & de Swart, Peter (eds.), Case,
valency and transitivity. Amsterdam & Philadelphia: J. Benjamins (Studies in Language Companion
Series, 77); 465-493.
Malchukov, Andrej L. & Haspelmath, Martin & Comrie, Bernard (eds.) 2010, Studies in ditransitive
constructions. A comparative handbook. Berlin & New York: W. de Gruyter.
Martínez Corripio, Israel & Maldonado, Ricardo 2010, "Middles and reflexives in Yucatec Maya. Trusting
speaker intuition" Berez, Andrea L. & Mulder, Jean & Rosenblum, Daisy (eds.), Fieldwork and
linguistic analysis in indigenous languages of the Americas. Honolulu: University of Hawai'i Press
(Language Documentation and Conservation, Special Publication, 2); 147-171.
Paris, Catherine 1985. “Relations actancielles et valence verbale en avar: effacement de l’actant et mise au
jour du sens”. Actances 1: 135-153.
Sullivan, Paul R. 1984, “Noun incorporation in Yucatec Maya." Anthropological Linguistics 26(2):138-160.
Verhoeven, Elisabeth 2007, Experiential constructions in Yucatec Maya. A typologically based analysis of a
functional domain in a Mayan language. Amsterdam & Philadelphia: J. Benjamins (Studies in Language
Companion Series, 88).
... This distribution is not accidental, and reflects the restrictions on the use of causatives and applicatives also observed elsewhere. In particular, (morphological) causatives are restricted to intransitives (Nedjalkov & Sil'nitskij 1969), and in some languages (like Yucatec Maya) are mostly restricted to inactive intrasitives (unaccusatives) (Lehmann 2015). On the other hand, applicatives show an opposite pattern, being infelicitous with unaccusative verbs. ...
... In particular, in some cases, the applicative sense seems to be primary, while the causative is a derivative. Consider Yucatec Maya featuring an "extraversive" transitivizer, which can be regarded as a lexical counterpart to the applicative derivation (Lehmann 2015). In most cases extraversive routinely applies to active intransitives introducing a new object (a path in (25b)), and is different from a causative (as in (25c)): ...
... (25) Yucatec Maya (Lehmann 2015(Lehmann : 1457 a. Le tsíimin-oʹ áalkab-nah tiʹ le beh-oʹ. dem horse-d2 run-compl loc dem path-D2 'The horse ran along the path. ...
This article explores the connection between expression of valency alternations and the overall morphological typology of a language from a cross-linguistic perspective. On the basis of a typological survey of empirical data in 40 geographically and genealogically diverse languages, it finds two universal tendencies relating to the scale of morphological types from most to least bound: fusional – agglutinative – isolating. First, the compatibility of morphological techniques used to express valency alternations does not extend further left than the overall morphological typology of the language. Second, it may extend further right, with extensive attestation of this possibility. The morphological expression of valency alternations is thus constrained by the overall morphology of the language, but tends to be pushed further towards the right.
Full-text available
In this article the author investigates the function of noun incorporation within the functional domain of participation, with a focus on the accommodation of peripheral participants by incorporation. Like most other languages, in Yucatec Maya the verb incorporates with preference participants at the lower end of the control hierarchy, such as the patient, the instrument, and, less frequently, the location. This yields an intransitive verb which is normally not used as such, but immediately re-transitivized in order to accommodate another, more peripheral participant. The latter may be an adjunct, a possessive complement, or a participant that did not even appear in the base situation. The analysis of syntactic relations in these constructions reveals an advanced degree of lexicalization of many incorporative verbs, which renders a purely syntactic analysis insufficient. On the other hand, noun incorporation in Yucatec covers a rather large area between verbal syntax and lexicon, which confers it a central position in the linguistic system.
Full-text available
Abstract The purpose of a classification of predicates is to provide the basisfor any explanations of different grammatical behavior of predicates that may be grounded in differences of lexical semantics. The scope of such a classificationhas to comprise not only verbs, but all words that can function as predicates, because the criteria must not be structural. The criteria of such a classificationare therefore such as characterize situations, participants and their roles in situations. The features form a hierarchy and are defined operationally
Causativization has a position in an intricate network of lexical relations and valency operations, including deagentive and passive, whose general function it is to express a given predicate in construction with different constellations of central participants. The corresponding tasks can be solved by (1) coding the causative or non-causative relation of an actant to the rest of the clause, or (2) leaving it to inference. In the first case, the relation may be coded (1.1) on the verb or (1.2) on the actants. If the relation is coded on the verb, this may be done (1.1.1) at the lexical level, by a paradigm of converse verb stems, (1.1.2) at the derivational level by deriving a verb stem with a different valency, or (1.1.3) at the syntactic level by a construction involving another verb. In the cases 1.1, the causative construction may be more marked compared with a base of lesser valency, or else the deagentive construction may be more marked compared with a base of higher valency. In case 1.2, a given stem is converted into a different valency frame according to the syntactic environment. The causative constellation may then be coded by a special case either on the causer or on the causee. Latin may be characterized as a language that makes very little use of the strategy that is cross-linguistically the most common one, viz. 1.1.2. Instead, it relies heavily on 1.1.1, 1.2 and 2. This is in consonance with its general aversion against derivational valency operations and its dependent-marking type.
Relationship terms with the usative suffix in Tzotzil and Yucatec Maya
  • Victoria R Bricker
Bricker, Victoria R. 1970, "Relationship terms with the usative suffix in Tzotzil and Yucatec Maya". Papers from the Regional Meeting of the Chicago Linguistic Society 6: 75-86.
Possession in Yucatec Maya. Second, revised edition
  • Lehmann
Lehmann, Christian 2002, Possession in Yucatec Maya. Second, revised edition. Erfurt: Seminar für Sprachwissenschaft der Universität (ASSidUE, 10).
Relations actancielles et valence verbale en avar: effacement de l'actant et mise au jour du sens
  • Catherine Paris
Paris, Catherine 1985. "Relations actancielles et valence verbale en avar: effacement de l'actant et mise au jour du sens". Actances 1: 135-153.
Noun incorporation in Yucatec Maya
  • Paul R Sullivan
Sullivan, Paul R. 1984, "Noun incorporation in Yucatec Maya." Anthropological Linguistics 26(2):138-160.