Content uploaded by Christian Lehmann
Author content
All content in this area was uploaded by Christian Lehmann on May 31, 2016
Content may be subject to copyright.
CLIPP
Christiani Lehmanni inedita, publicanda, publicata
titulus
(Coauctrice Elisabeth Verhoeven:)
Extraversive transitivization in Yucatec Maya and
the nature of the applicative
huius textus situs retis mundialis
http://www.christianlehmann.de/publ/
lehmann_verhoeven_extraversion.pdf
dies manuscripti postremum modificati
25.08.2005
occasio orationis habitae
Workshop ‘Case, valency and transitivity’, Universiteit
Nijmegen, June 17-20, 2003 (ab Elisabeth Verhoeven)
volumen publicationem continens
Kulikov, Leonid & Malchukov, Andrej & de Swart, Peter
(eds.), Case, valency and transitivity. Amsterdam &
Philadelphia: J. Benjamins (SLCS, 77).
annus publicationis
2006
paginae
465-493
Extraversive transitivization in Yucatec Maya
and the nature of the applicative
Christian Lehmann & Elisabeth Verhoeven
(Universities of Erfurt and Bremen)
Abstract
Just as actor-focused transitivization may essentially be equated with cau-
sativization, undergoer-focused transitivization is often equated with ap-
plicative formation. Transitivization in Yucatec Maya by means of the
morpheme -t resembles applicative formation in other languages to some
extent. However, it differs from the latter in being basically a lexical op-
eration with only limited syntactic regularity.
Yucatec Mayan transitivization by the suffix -t is described and analyzed
with the aim of refining the concept of applicative. Special attention is
given to a possible functional transition between plain undergoer-focused
transitivization (named ‘extraversion’) and applicative formation. Such a
transition is based on the kind of thematic roles typically involved in the
two constructions.
1
1. Introduction
The aim of this contribution is to assess the place of a particular transitivization proc-
ess of Yucatec Maya (YM), called extraversion, in the functional typology of partici-
pation, in particular of operations of installation and suppression of argument posi-
tions, of promotion and demotion of verbal dependents. We will compare extraversion
with current assumptions about applicative formation and try to show that while it
does render some of the service commonly attributed to applicatives, it is peculiar in
other ways and does not obey certain generalizations that have been made about ap-
plicatives. The solution that we propose is to restrict the concept of applicative to cer-
tain syntactically regular promotion processes and to distinguish it from the concept
of the extraversive, which is essentially a lexical-derivational process of providing an
intransitive base with an undergoer slot.
1
We thank Julia Galiamina and an anonymous reviewer for very helpful comments on an earlier ver-
sion.
Extraversive transitivization in Yucatec Maya
2
The paper is organized as follows: Section 2 gives an outline of the theoretical back-
ground and sets out the basic concepts of the analysis. Section 3 introduces extraver-
sion in YM, concentrating on the types of thematic roles affected by this process. It is
shown that while some cases of extraversion are indeed similar to what is commonly
called applicative formation, most cases do not afford the promotion associated with
the latter. Section 4 characterizes and defines the operations of applicative formation
and extraversion in more detail and delimits them against each other. It thus sets the
frame for a closer investigation of undergoer-focused transitivization in YM, which is
undertaken in section 5. Here, it is shown that the YM operation mainly sticks to the
lexical side of the continuum of undergoer-focused transitivization. Those cases that
are more productive and regular seem to be either a more recent development, as e.g.
addressee-applied-objects of communication verbs, or extraversion is combined with
incorporation, resembling thus the rearrangement type of applicative formation. Sec-
tion 6 posits a functional transition between extraversion and applicative formation.
2. Basic concepts
The analysis of the YM situation and its typological comparison will involve a few
concepts that have been used in different ways and which we therefore do better in de-
fining at the outset. A verb with its dependents designates a situation consisting of
participants assembled around an immaterial center called the situation core. Depend-
ing on the specificity of the selection restrictions, a certain kind of participant may be
more or less inherent in the concept of a predicate. Extreme cases of inherence are
provided by verbs like ‘dream’ and ‘dance’, whose second participant may be exteri-
orized in the form of a cognate object. The conceptual operation of exteriorization
2
has a counterpart in interiorization, which manifests itself in the incorporation of
nominal expressions in the verb.
Figure 1 serves to enumerate those thematic roles that will be taken up below and to
arrange them by the two most important parameters, involvement and con-
trol/affectedness.
2
Cf. Lehmann 1991, section 3.2 on exteriorization of participants. A holistic, undifferentiated situation
is semantically represented just by the situation core which contains the participants. Exteriorization of
a participant means that it receives its own linguistic representation which in turn comes along with in-
dividuation and referential independence. Thus, exteriorization can be conceived as an operation which
gradually brings participants (included in the situation core) to the fore and opposes them syntagmati-
cally to the situation core.
Christian Lehmann & Elisabeth Verhoeven 3
Figure 1. Involvement and control of participants
control affectedness
actor
undergoer
central agent force theme patient
experiencer
recipient/addressee/goal
emitter/source
involvement
beneficiary/place
peripheral comitative/instrument
A participant may be more or less intimately involved in a situation (cf. Lehmann,
Shin and Verhoeven 2000, section 2.3.1). A central participant is inherent in the con-
cept of the predicate, so that if it is subtracted, the concept of the predicate changes.
Peripheral participants do not concern the concept of the predicate, are compatible
with many different predicates and may be so remote from the situation core that they
may rather be part of another situation. Central participants are typically coded as
verb complements, peripheral participants are typically coded as dependents of addi-
tional relators which may be case relators or even (subordinate or ‘co-subordinate’)
verbs. The terms ‘predicate’ and ‘argument’ will be used to refer to the language-
specific semantic representation of a verb and its complement.
There are operations of moving a peripheral participant to the center or, conversely,
moving a participant out of the center of the situation. At the level of syntax, these
appear as operations of promotion and demotion of verbal dependents. These concepts
presuppose a hierarchy of syntactic functions that may roughly be depicted as in
Figure 2 (where, for the sake of simplicity, syntactic ergativity is ignored):
Figure 2. Hierarchy of syntactic functions
1 subject
2 direct object / primary object
3 indirect object / secondary object
4 other complement
5 adjunct
Any operation that assigns a verbal dependent a syntactic function higher up in Figure
2 is an operation of promotion; any operation that assigns it a function lower down is
an operation of demotion. If a participant is interiorized or not exteriorized, it has no
syntactic function. In section 4.3, we will come back to the issue of whether ‘no syn-
tactic function’ is the lowest position on Figure 2.
In Figure 1, the two most central participants are the agent who controls the situation,
and the patient who is affected by it. These two notions are schematized in the form of
Extraversive transitivization in Yucatec Maya
4
the two macroroles of ‘actor’ and ‘undergoer’.
3
The semantic valence of most bivalent
verbs uses the template of opposing an actor to an undergoer, i.e. a dependent that has
more vs. less control in the situation. Promoting a peripheral participant often
amounts to subsuming it under the undergoer macrorole. Many languages provide for
a third macrorole, the indirectus (cf. Lehmann, Shin and Verhoeven 2000), which es-
sentially neutralizes those roles in the center of Figure 1 that belong to highly em-
pathic participants that are neutral or ambivalent as to the control cline, such as recipi-
ent, addressee, experiencer and beneficiary.
There are essentially two motivations for promoting a dependent on Figure 2. It may
either be motivated semantically as drawing it into the control cline and subsuming it
under one of the macroroles of actor, undergoer or indirectus. Or else the promotion
may be motivated by functional sentence perspective (alias information structure),
more particularly as granting topic continuity to the participant in question. For in-
stance, given the context She led him to her desk and __, (1a) (without the repeated
she) is a more likely continuation than (1b).
(1) a. She showed him a book.
b. She showed a book to him.
Operations of promotion and demotion lead to a rearrangement of the syntactic struc-
ture of the clause, in other words to a paradigmatic relationship between two syntactic
constructions. For the morphological structure of the verb whose dependents are con-
cerned, this may mean either of two things:
• either a deverbal verb derivation transfers the stem into a different valence class
and marks this by some morphological process on the verb stem,
• or the same verb stem is used in two distinct valence frames, which may be de-
scribed as (valence) conversion of the stem.
We will keep these two processes apart, i.e. we will not consider conversion as a kind
of “zero-marked” derivation.
4
In the case of a derivation, it is generally easy to recog-
nize its direction, i.e. to tell which stem is the base and which is derived. In the case of
conversion, the verb stem itself is not affected. Of the two syntactic constructions, one
may be simpler than the other, i.e. involve syntactic functions higher up on Figure 2
or use less marking by case relators; and this may then be considered basic. By this
criterion, we may say that in the English dative shift illustrated in (1), the direction of
conversion is from a to b. There may be other criteria such as increased constraints on
the distribution of the converted version, which we may forego here. Sometimes, no
direction of conversion may be discerned, in which case we simply recognize cate-
gorical indeterminacy for such a verb stem.
3
For the concept of macroroles, see Foley and Van Valin 1984, Van Valin and LaPolla 1997 and Leh-
mann, Shin and Verhoeven 2000.
4
We are thereby radicalizing the position taken in Dixon and Aikhenvald 1997, 2000 and Peterson
1999, for which derivational marking of the applicative is only the prototypical case.
Christian Lehmann & Elisabeth Verhoeven 5
The syntactic or derivational operations involved may be viewed as installing or
blocking a valence position – a “slot” – on a verb. The two slots most commonly af-
fected by such operations are the ones associated with the actor and the undergoer
macroroles. Those are then transitivity changing operations, which may be classified
as in Table 1.
Table 1. Transitivity-changing operations
macrorole
operation actor undergoer
installation
actor-focused
transitivization:
causative
undergoer-focused
transitivization:
applicative, extraversive
suppression
actor-focused
detransitivization:
passive, anticausative
undergoer-focused
detransitivization:
antipassive, introversive
We have not used the terms ‘valence increase’ and ‘valence reduction’ to label the
row entries of Table 1. First, quantitative valence – the number of argument slots of a
verb – is not what is at stake here. Very often, it remains unchanged after such an op-
eration. Especially, the installation of a valence slot for a new argument is often at the
expense of the demotion or suppression of an argument provided in the base. What
these operations aim at is either to accommodate a certain argument in verb valence
or, on the contrary, to eliminate it from verb valence. Augmentation or reduction of
the number of slots may be a contingent consequence. Furthermore, valence is a prop-
erty of a verb stem, not of a verb form. Therefore, inflectional operations such as
SAE-style passive or Australia-style antipassive are not, strictly speaking, valence-
changing operations.
On the other hand, these operations are sometimes called ‘(direct) object addi-
tion/deletion’ etc. (e.g. Dixon and Aikhenvald 1997, Haspelmath and Müller-Bardey
2004). However, whether a noun phrase (in a certain syntactic function) is mentioned
or not mentioned is one matter; and the manipulation of valence slots of a verb is an-
other matter. Grammatical and derivational operations are essentially concerned with
the latter.
We intend our conception to be applied generally to languages with different align-
ment types of fundamental relations. Therefore we have resorted to the terms ‘actor’
and ‘undergoer’, because these are neutral as to syntactic ergativity or even active-
inactive clause structure. However, in what follows, only accusative clause structures
will be considered, so we may as well speak of ‘(transitive) subject’ and ‘direct ob-
ject’ instead.
Here we are not concerned with the ‘actor’ column of Table 1. The two operations
affecting the undergoer slot may be exemplified from YM:
Extraversive transitivization in Yucatec Maya
6
(2) a. hun-túul máak túun yáakan
one-CL.AN person PROG:SBJ.3 groan
mèen hach yah ba’x k yùuchul ti’
because very ache what IMPF SBJ.3:happen:INCMPL him
‘a person groans because it hurts him what is happening to him’
(MPK_016)
b. tu yáakan-t-ah u yahilo’
PRV:SBJ.3 groan-TRR-CMPL POSS.3 ache:REL:D2
‘he bemoaned his pain’ (ACC_0463)
(3) a. tin bul-ah tuláakal in tàak’in
PRV:SBJ.1.SG gamble-CMPL all POSS.1.SG money
‘I gambled away all my money’ (RMC_0231)
b. ko’ne’x bùul
lets.go gamble\INTRV
‘let’s play cards’ (RMC_0234)
The terms ‘extraversion’ and ‘introversion’ (introduced in Paris 1985: 145-146) refer
to derivational operations on a verb stem, thus lexical operations that change the syn-
tactic potential of that stem. The main verb in (2a) is intransitive; in (2b), a direct ob-
ject slot is installed on it by extraversion, the derivational process that is the object of
this study. The main verb in (3a) is transitive; in (3b) its direct object slot is blocked
by introversion, whose morphological mark is low tone in the base. The latter process
will briefly be resumed in section 4.2.
3. Extraversion in Yucatec Maya
3.1. Basic facts
YM has actor-focused transitivization by means of a causative suffix whose main al-
lomorph is -s, and undergoer-focused transitivization by means of the suffix -t
(glossed as
TRR
‘transitivizer’ throughout). The latter is a very common and frequent
process in YM grammar (cf. Lehmann 1993, 2002, ch. 2.4.3, 5.3; Bohnemeyer 2004).
Most commonly, it works on an intransitive verbal base. The action denoted by the
verb is thus extended to an undergoer which is affected by it. Verbal bases that take
the suffix -t typically belong to the active class of intransitive verbs. Suffixation of -t
results in a recategorization so that the derived verb belongs to the class of transitive
verbs. This was already illustrated in (2) above and is schematized in Figure 3.
Figure 3. Extraversion in Yucatec Maya
[ [ X ]
V-intr
–t ]
V-tr
action
Christian Lehmann & Elisabeth Verhoeven 7
YM has a set of more than 150 (mostly verbal) roots which can be marked by the suf-
fix -t. Furthermore -t is productive and obligatory in transitive compound and incorpo-
rative verbs (cf. Owen 1973), which we will come back to in section 5.5. Moreover,
all transitive verb stems based on Spanish loans, like k-in formar-t-ik (
IMPF
-
SBJ
.1.
SG
form-
TRR
-
INCMPL
) ‘I form it’, bear a final -t. Next to verbs, nouns and, very rarely,
adjectives occur as bases. Thus, this process is clearly defined by its output, which is a
transitive verb stem, whereas it is more liberal with respect to input categories.
In contrast to certain other languages that have a single generic transitivization proc-
ess, YM generally keeps actor-focused and undergoer-focused transitivization apart.
One exception is constituted by some active intransitive verbs of sound emission and
manner of motion that derive a causative verb with -t (e.g. tirix ~ tirix-t ‘rattle ~ make
rattle’, balak’ ~ balak’-t ‘roll ~ roll sth.’, cf. Bohnemeyer 2004). Factitive verbs are a
further exception: on such deadjectival transitive verb stems as chak-kun-t (red-
FACT
-
TRR
) ‘redden’, -s and -t appear to be in free variation with most such verbs for most
speakers. However, derivation in -t is the only process of undergoer-focused transi-
tivization that the language has. In particular, there is no contrast among derivational
suffixes to disambiguate the thematic role of the direct object. Thus, -t is a rather gen-
eral marker of transitivity.
3.2. Peripheral thematic roles
Transitivization is commonly analyzed by seeking a transformational relationship be-
tween the transitive construction and an intransitive base version such that the partici-
pant coded as the direct object of the former corresponds to some adjunct of the latter.
This adjunct is typically adjoined by some suitable case relator. In YM, this could be
one of a set of prepositions. We will first review a set of verbs in which such a regular
syntactic relationship between an intransitive base verb and its transitivized counter-
part does work out. In (4a) the experiential stimulus u na’ ‘his mother’ is joined to the
intransitive verb ts’íikil ‘feel angry’ by the generic preposition ti’ (
LOC
) ‘at, to’ etc.,
while in the transitivized version of (4b) it takes the function of the direct object. (5)
illustrates a similar alternation with a local participant.
(4) a. táan u ts’íikil (ti’ u na’)
PROG SBJ.3 feel.angry LOC POSS.3 mother
‘he is annoyed (with his mother)’ (HNAZ_0019.01)
b. táan u ts’íikil-t-ik u na’
PROG SBJ.3.SG feel.angry-TRR-INCMPL POSS.3.SG mother
‘he is annoyed with / is scolding his mother’ (NMP_0362)
Extraversive transitivization in Yucatec Maya
8
(5) a. táan u bin bàab (ich le ha’-o’)
PROG SBJ.3 go swim in DEF water-D2
b. táan u bin u bàab-t le ha’-o’
PROG SBJ.3 go SBJ.3 swim-TRR(SUBJ) DEF water-D2
‘he is going to swim (in the water)’ (MPK_018/EMB)
These examples resemble intransitive-based applicatives in other languages: the ap-
plied objects of the b-versions are prepositional adjuncts in the intransitive frame of
the a-versions.
We now apply the same method to various groups of extraversive verbs. That is, we
try to paraphrase the transitive version with a construction based on the intransitive
version, trying out such prepositions that promise to be viable, i.e. to lead to a con-
struction that is grammatical, synonymous with the transitive version and in a regular
(“transformational”) relationship with it. The verbs will be grouped by the thematic
role mapped onto the direct object function of the extraversive verb. In the present
section, we consider various peripheral participants; in the next section, we apply the
method to central participants. We start with (6), where a verb of emotional expres-
sion is directed towards its metaphorical goal, coded as a direct object in (6b).
(6) a. láahk’ìin táan u che’h
all:day PROG SBJ.3 laugh
‘he laughs the whole day’ (CPP_0018)
b. t-in che’h-t-ah in wíits’in
PRV-SBJ.1.SG laugh-TRR-CMPL POSS.1.SG younger.sibling
‘I laughed at / derided my younger sibling’ (AVC_0031)
c. *h che’h-nah-en ti’ in wíits’in
PRV laugh-CMPL-ABS.1.SG LOC POSS.1.SG younger.sibling
intended: ‘I laughed at/about my younger sibling’ (ACC)
d. h che’h-nah-en yéetel/yóosal in wíits’in
PRV laugh-CMPL-ABS.1.SG with/because.of POSS.1.SG younger.sibling
‘I laughed with/because of my younger sibling’ (AVC_0033)
(6c) and (6d) represent various attempts at accommodating that participant in the in-
transitive base frame of (6a). (6c) uses the preposition ti’ which worked for (4); but it
is ungrammatical. (6d) uses other prepositions; but then the participant in question
clearly bears different roles. Similar verbs include òok’(ol)(-t) ‘cry, weep ~ mourn’,
áakan(t) ‘groan, complain’ (cf. (2) above), sunkal(-t) ‘grunt; roar, bawl’, héenkal(-t)
‘grunt, roar (of wild animals)’, etc.
With another subgroup of transitivized verbs, the direct object represents the ad-
dressee. In (7b) the verb xóob ‘whistle’ is transitivized and in this way directed to-
wards the addressee.
Christian Lehmann & Elisabeth Verhoeven 9
(7) a. le xibpal-o’ túun xóob
DEF man:child-D2 PROG:SBJ.3 whistle
‘the boy is whistling’ (ACC)
b. le xibpal-o’ túun xóob-t-ik
DEF man:child-D2 PROG:SBJ.3 whistle-TRR-INCMPL
le chàan x ch’úuppal-o’
DEF little F woman:child-D2
‘the boy is whistling at the girl’ (AME_0052)
c. *le xibpal-o’
DEF boy-D2
túun xóob ti’ le chàan x ch’úuppal-o’
PROG:SBJ.3 whistle LOC DEF little F woman:child-D2
intended: ‘the boy is whistling at the girl’ (ACC)
As before, (7c) represents a futile attempt at accommodating the same participant in
the intransitive base frame of (7a).
A further subgroup includes experiential verbs which, if transitivized, code the
stimulus in direct object function. In (8), the experiential verb tùukul ‘think’ is
directed toward the stimulus in na’ ‘my mother’.
(8) a. táan in tùukul
PROG SBJ.1 think
‘I am thinking’ (ACC)
b. táan in tùukult-ik in na’
PROG SBJ.1 think:TRR-INCMPL POSS.1SG mother
‘I miss my mother’ (ACC)
c. *táan in tùukul ti’ in na’
PROG SBJ.1.SG think LOC POSS.3 mother
intended: ‘I am thinking about / missing my mother.’ (ACC)
And again, there is no way of expressing the same thematic role with the intransitive
verb. Similar cases are náay(-t) and wayáak’(-t), both ‘dream of/about’, cha’n(-t)
‘contemplate, look at, enjoy seeing’, kanáan(-t) ‘watch over’, etc.
With a further set of YM verbs, prepositional objects in intransitive frames do seem to
alternate with direct objects in the transitivized frame, but they represent different
thematic roles.
5
For example, with the motion verb síit’ ‘jump’ the direct object refers
to the traversed entity, as in (9a). (9b) shows that this participant cannot be joined as
an adjunct to the respective intransitive verb. Instead, a complex sentence has to be
5
This does not, however, imply that a given verb may take only one type of participant as direct object.
As with basic transitive verbs, the semantics of the direct object may be very general covering all kinds
of undergoers. In certain cases, the concrete role depends on the empathy of the participant in question
and is additionally inferred from the situation denoted and from the general context.
Extraversive transitivization in Yucatec Maya
10
formed which explicates the role of the participant in question. Joining a prepositional
phrase with óok’ol ‘on/above/over’ directly to the intransitive verb results in a change
in meaning: in (9c), the object is not traversed, but serves as the support of the action.
(9) a. táan in sen síit’-t-ik le sùum-a’
PROG SBJ.1.SG very jump-TRR-CMPL DEF rope-D1
‘I am (perpetually / really) jumping over this rope’ (AVC_0038)
b. h síit’-nah-en
PRV jump-CMPL-ABS.1.SG
káa h máan-en yóok’ol le kòot-o’
CNJ PRV pass(CMPL)-ABS.1.SG on DEF wall-D2
‘I jumped over the wall’ (AVC_0037)
c. h síit’-nah-en yóok’ol le kòot-o’
PRV jump-CMPL-ABS.1.SG on DEF wall-D2
‘I jumped (being) on the wall’ (AVC_0036)
Other verbs that adjoin different undergoer roles in their intransitive vs. transitive
frames are báaxal ~ báaxt ‘play’, ba’te’l(-t) ‘fight’, meyah(-t) ‘work’, náahal(-t)
‘earn’, pàakat ~ pakt ‘look at, gaze’, cha’n(-t) ‘contemplate, look at, enjoy seeing’,
ch’èeneb(-t) ‘peek, peer, spy’, kanáan(-t) ‘watch over’, óok’ot(-t) ‘dance’ etc.
The upshot of this section is that while the direct object of an extraversive verb may
code a variety of participants, only exceptionally may the same participant be coded
as an adjunct of the intransitive base.
3.3. Central thematic roles
We now move on to such transitivized verbs whose direct object plays a central the-
matic role such as patient or theme. YM lexicalizes a number of action concepts in-
volving a patient like ‘write’, ‘sweep’, ‘weed’, ‘shell’ (some of which are labile in
English or German) by intransitive verbs, adding the transitive marker when the verb
is used with an object. This group of verbs includes mahàan(-t) ‘borrow, lend’, páay(-
t) ‘haul water, pull on a rope’, ya’ch’(-t) ‘dissolve’, tsi’k(-t) ‘shred’, cháal(-t) ‘lute’,
máay(-t) ‘strain’, húuy(-t) ‘stir’, tsíik(-t) ‘comb’, pak’ach(-t) ‘make tortilla’, sakal(-t)
‘weave’, lòobil(-t) ‘fight, slap’, wáay(-t) ‘bewitch, put a spell on’, etc. Their use is il-
lustrated by (10).
(10) a. Húuy-t le sa’-o’ bik táak’-ak!
stir-TRR(IMP) DEF atole-D2 PROHIB stick\DEAG-SUBJ
‘Stir the atole lest it sticks!’ (ACC_0265)
b. h húuy-nah-en
PRV stir-CMPL-ABS.1.SG
‘I stirred (sth.)’
Christian Lehmann & Elisabeth Verhoeven 11
c. *Húuy-nen ti’ / ich le sa’-o’!
stir-ITR.IMP LOC / in DEF atole-D2
intended: ‘Stir (in) the atole!’ (ACC)
Again, the transitivized versions of tsikbal(-t) ‘converse, talk about’, nu’k(-t) ‘ex-
plain’, tse’k(-t) ‘preach, lecture, advise, scold’, p’a’s(-t) ‘
mock, criticize, ridicule’,
xíix(-
t) ‘sift through, cull’, éets’(-t) ‘imitate’, se’n(-t) ‘cough’, síin(-t) ‘blow nose’, etc. take
a theme argument as direct object, as illustrated in (11).
(11) a. t-in p’a’s-t-ah le ba’x t-u mèet-ah-o’
PRV-SBJ.1.SG mock-TRR-CMPL DEF thing PRV-SBJ.3 do-CMPL-D2
‘I mocked / criticized the thing he did’ (RMC_1073)
b. h p’àa’s-nah-en
PRV mock-CMPL-ABS.1.SG
‘I mocked (sth./sb.)’
c. *h p’àa’s-nah-en
PRV mock-CMPL-ABS.1.SG
ti’ / yéetel le ba’x t-u mèet-ah-o’
LOC / with DEF thing PRV-SBJ.3 do-CMPL-D2
intended: ‘I mocked / criticized the thing he did’ (ACC)
These two sets of verbs share with the ones represented by (6) – (9) the fact that the
participant figuring as direct object of the transitive verb cannot be adjoined to the in-
transitive verb. The sets of verbs of this and of the previous section differ in the cen-
trality of the participant coded as direct object. However, the examples show that this
semantic difference is only partly correlated with different structural behavior: while
some peripheral participants can be adjoined to the intransitive base verb, no central
participant ever can. The latter is, in fact, less surprising, because a participant that
could be adjoined in a regular way by a preposition would probably not be a central
participant.
The data of this section shows that YM lexicalizes as intransitive verbs not only ver-
bal concepts that range in the central region of the effectiveness continuum
6
as e.g.
‘look at’, ‘laugh at’, etc., but also concepts with a higher effectiveness value, e.g. the
equivalents of ‘lend’, ‘shred’, ‘strain’, stir’, ‘shell’ and others. Both may have their
origin in two classes of nouns common to most Mayan languages. The first of these
comprises action concepts like ts’íib ‘writing, write’, meyah ‘work, worker’, tsikbal
‘chat, chatting, story’ etc. (cf. Kaufman 1990). The second class comprises concrete
nouns (denoting physical objects) such as oxo’m ‘shelled corn’ – oxo’m-t ‘shell’,
pak’ach ‘tortilla’ – pak’ach-t ‘make tortilla’, sakal ‘cloth’ – sakal-t ‘weave’. Accord-
ing to Bohnemeyer 2002a: 179, 2002b, both classes of nouns would have been recate-
gorized as active intransitive verbs in YM or in the Yucatecan branch.
6
See Tsunoda 1981, and cf. the concept of ‘transitivity’ in Hopper and Thompson 1980.
Extraversive transitivization in Yucatec Maya
12
To summarize section 3, we may say that YM extraversion shares with applicative
formation of other languages its basic nature of being an undergoer-focused transitivi-
zation process. It also sometimes behaves specifically like applicative formation in
taking part in a transformational relationship between two constructions, one in which
a certain participant is coded as an adjunct of an intransitive base and another in
which the same participant is coded as the direct object of the transitivized base. This
is, however, not typical of the YM extraversive. In general, the intransitive base is not
used to speak about that participant that appears with the extraversive. Therefore we
do not subsume the YM extraversive under the typological concept of the applicative,
but instead compare them in the following section.
4. Types of undergoer-focused transitivization
In Table 1, we have introduced the applicative and the extraversive as two kinds of
undergoer-focused transitivization. The distinction intended thereby is not generally
made in the literature. Typically, the term ‘applicative (formation)’ is used for what
we call undergoer-focused transitivization. Here are two representative examples:
Corresponding to the first row of our Table 1, Dixon and Aikhenvald 2000 have the
following subdivision: “5. Valency increase … (1) Causative … (2) Applicative”.
Haspelmath and Müller-Bardey 2004 have the section headings: “3. Valency-
increasing categories … 3.1. Object-adding categories: the applicative … 3.2. Agent-
adding categories: the causative”. This concept commonly embodies the following
two suppositions although these are not necessarily stated explicitly: First, although
the processes in question are derivational processes, they can be stated in syntactic
terms, i.e. a version of the transformational approach that we have applied in section 3
is taken for granted. Secondly and more specifically, installing a direct object slot on a
verb amounts to a promotion of one of its dependents. This is stated explicitly in
Dixon and Aikhenvald 2000: 14: “Applicative derivations all have a common syntac-
tic effect, with a peripheral participant being brought into O function …”.
Occasionally, an even broader concept of the applicative is found. Peterson’s (1999)
concept of applicative involves coding a “semantically peripheral object in a more
central morphosyntactic (and sometimes discourse) way than would otherwise be ex-
pected for it”. This includes promotion to the indirect or secondary object slot, thus
presumably any upward movement in Figure 2 that does not reach level 1. Haspel-
math and Müller-Bardey (2004: 1136), too, speak of “dative-adding applicatives”.
In what follows, we shall argue for a narrower concept of applicative which is op-
posed to extraversive. In characterizing the two concepts, we consider both formal
and functional criteria.
7
We begin by discussing applicative formation (section 4.1)
7
Cf. Comrie 1985, Drossard 1991, Lehmann 1991, Dixon and Aikhenvald 1997, 2000, Peterson 1999,
Haspelmath and Müller-Bardey 2004.
Christian Lehmann & Elisabeth Verhoeven 13
followed by extraversion (section 4.2) before arranging both processes on a contin-
uum of undergoer-focused transitivization in section 4.3.
4.1. Applicative formation
4.1.1. Applicative as a promotion process
The simplest form of applicative is found with intransitive bases, as in (12) from In-
donesian (Shibatani 1996: 159).
(12) a. Saya duduk di kursi.
I sit in chair
‘I am sitting on the chair.’
b. Saya men-duduk-i kursi.
I ACT-sit-APPL chair
‘I am occupying the chair.’
Here the adjunct of the intransitive version (12a) is promoted to direct object in the
applicative transitive version b.
If the base is already transitive, as in (13), then the first phase of the process is the
same as before. This, however, ousts the direct object of (13a) from its position; so in
a second phase, this is demoted onto some lower level of Figure 2, in the present case,
into the function of a secondary object.
(13) a. Saya akan mem-beli buku untuk orang itu.
I FUT ACT-buy book for person DEF
‘I will buy a book for the man.’
b. Saya akan mem-beli-kan orang itu buku.
I FUT ACT-buy-APPL person DEF book
‘I will buy the man a book.’
German does not have a secondary object function. Here the demoted direct object
ends up as an adjunct, as in (14) (from Comrie 1985: 313f).
(14) a. Hans pflanzt [Bäume]
DO
[im Garten]
PO
.
‘Hans plants trees in the garden.’
b. Hans bepflanzt [den Garten]
DO
[mit Bäumen]
PO
.
‘Hans plants the garden with trees.’
The German derivation by the prefix be- may promote to direct object function not
only peripheral dependents, as in (14), but also indirect objects, as in (15).
(15) a. Maria schenkte [dem Mann]
IO
[ein Buch]
DO
.
‘Mary gave a book to the man (as a present).’
Extraversive transitivization in Yucatec Maya
14
b. Maria beschenkte [den Mann]
DO
[mit einem Buch]
PO
.
‘Mary presented the man with a book.’
Various Bantu languages including Swahili, Chichewa and ChiMwi:ni possess an ap-
plicative operation of the same type. (16) from Chichewa (Baker 1988: 229) features
the recipient as a prepositional object in the a-version. In the applicative b-version it is
promoted to direct object, while the erstwhile direct object is demoted to secondary
object.
(16) a. Mbidzi zi-na-perek-a msampha kwa nkhandwe.
zebras SBJ-PST-hand-FV trap to fox
‘The zebras handed the trap to the fox.’
b. Mbidzi zi-na-perek-er-a nkhandwe msampha.
zebras SBJ-PST-hand-APPL-FV fox trap
‘The zebras handed the fox the trap.’
Thus, irrespective of the transitivity of the base verb and the resulting valence of the
derived verb, applicative formation presupposes the hierarchy of syntactic functions
introduced in Figure 2 and involves a promotion to direct or primary object function
as visualized in Figure 4.
Figure 4. Applicative formation
1 subject
2 direct object / primary object
3 indirect object / secondary object
4 other complement
5 adjunct
4.1.2. Functions of the applicative
The main function of applicative formation is the centralization of the participant con-
cerned. In semantic terms, this entails an increase in involvement and affectedness,
where these concepts are understood as visualized in Figure 1. In the German (17a),
the burglar is involved more indirectly and coded as an indirect object, while in (17a’)
he is directly involved in the situation and coded as a direct object. The verb is transi-
tivized by the prefix ver-. A similar functional distinction of involvement is conveyed
by the prepositional vs. direct object marking in (17b)/(17b’), featuring again the ap-
plicative prefix be- (cf. (14)).
(17) a. Paul folgte dem Einbrecher.
‘Paul followed the burglar.’
a’. Paul verfolgte den Einbrecher.
‘Paul pursued the burglar.’
Christian Lehmann & Elisabeth Verhoeven 15
b. Paul herrschte über die Teutonen.
‘Paul reigned over the Teutons.’
b’. Paul beherrschte die Teutonen.
‘Paul governed/controlled the Teutons.’ (Lehmann 1991: 207-8)
Note that the examples in (17) differ from those of (14) and (15) regarding the transi-
tivity of the base verb. However, irrespectively of whether the promotion involved in
applicative formation occasions the demotion of an argument occupying the direct ob-
ject function of the base verb, as in (14) and (15), or the direct object function is
newly installed on the verb, as in (17), applicative formation principally aims at
heightening the involvement of the promoted participant.
In (14) and (17), there is not only an increase in involvement of a peripheral partici-
pant but also an increase in its affectedness in the situation. In the basic versions, the
participant in question is only weakly or partly affected. Through the operation of ap-
plicative formation it becomes strongly or totally affected. In the a-version of (14), for
instance, the locative participant may only be partly affected, the trees occupying only
part of the garden, while in the applicative version (14b), the whole garden is planted
with trees (cf. Comrie 1985: 314).
8
From the point of view of information structure, topic continuity plays an important
role in the foregrounding of a participant (cf. Givón 1983). Given that topic continuity
requires topical participants to be in higher-ranked syntactic functions and applicative
formation enables a peripheral participant to appear in just such a function, it is often
performed on a peripheral participant that is repeated or topicalized (cf. Rude 1986,
Peterson 1999, ch. 3). The examples in (18) and (19) show a participant that is topical
in the first clause and then referred to again in the second clause. The relative pro-
nouns in (18a) and (19a) are in direct object function with respect to the subordinate
applicative verbs, while in (18b) and (19b) they depend on appropriate local preposi-
tions. From the point of view of functional sentence perspective, (18a) is better than
(18b); and the same goes for (19a) and (19b).
(18) a. das Appartment, das Nicole besitzt und das Chantal bewohnt
‘the apartment that Nicole owns and Chantal inhabits’
b. das Appartment, das Nicole besitzt und in dem Chantal wohnt
‘the apartment that Nicole owns and that Chantal lives in’
(19) a. der Bürgersteig, der vor unserem Haus ist und den er (mit Sand)
(be)streuen muß
‘the pavement that is in front of our house and that he has to strew (with
sand).’
8
Cf. furthermore Michaelis and Ruppenhofer (2001, ch. 5.3.3), who identify an intensification of the
action after applicative derivation with be- in examples like (17b) and other pairs such as schimpfen ~
beschimpfen ‘scold at ~ insult’, lehren ~ belehren ‘teach ~ instruct, inform’ etc.
Extraversive transitivization in Yucatec Maya
16
b. der Bürgersteig, der vor unserem Haus ist und auf den er (Sand) streuen
muß
‘the pavement that is in front of our house and that he has to strew (sand)
on.’
Thus, promotion of a peripheral participant in applicative formation is motivated by
goals concerning thematic roles and/or information structure.
4.2. Extraversion
Extraversion is well attested in those languages (e.g. Oceanic languages) whose basic
verbs are usually intransitive and which mark the verb as transitive when it is used
with a direct object. In this case, there is generally no regular alternative that would
allow accommodating the direct object participant of the derived transitive verb in the
‘underlying’ intransitive frame. One such Oceanic language is Tolai, illustrated by
(20):
(20) a. A vavina i momo.
ART woman SBJ.3 drink
‘The woman drank (something).’
b. A vavina i mom-e ra tava.
ART woman SBJ.3 drink-TRR ART water
‘The woman drank the water.’ (Mosel 1991: 248)
(20) shows a derivation of an intransitive base verb with the aim of joining a further
‘new’ participant in direct object function, thus extraversion as in YM. The semantics
of the process consists of directing the action denoted by the verb towards a further
participant. The participant in question is typically intrinsic to the situation denoted by
the base verb, as the patient in (20).
9
In (20b) the participant is exteriorized, i.e. it re-
ceives its own lexical representation.
It appears that languages have a choice as to the valence pattern they use in the lexi-
calization of action concepts involving such an intrinsic undergoer.
10
One alternative
is to lexicalize such concepts as basic intransitive verbs and to apply extraversion if
the undergoer must be exteriorized. This is illustrated by (20) and equally by the YM
verb for ‘eat’: hàan is basically intransitive and has to be transitivized if the thing
eaten is to be joined, viz. hàan-t ‘eat sth.’. The other alternative is to lexicalize such
concepts as basic transitive verbs and to apply introversion – the mirror image of ex-
traversion – to focus on the action as such. This is shown in (21) from YM with the
9
Cf. the discussion of omitted objects in English in Rice 1987, section 5.4, where it is argued that they
are present on a conceptual level.
10
There is an analogous alternative concerning the provision of an actor with process concepts such as
‘break’. Such verbs may either be basically intransitive and be transitivized by causativization, or they
may be basically transitive and be detransitivized by anticausativization. Cf. Haspelmath 1993 and
Nichols, Peterson and Barnes 2004.
Christian Lehmann & Elisabeth Verhoeven 17
basic transitive verb kon ‘sell’, which is detransitivized in the b-version by introver-
sion.
(21) a. Úuchak wáah a kon-ik to’n le kùuts-o’?
possible INT SBJ.2 sell-INCMPL us DEF pheasant-D2
‘Is it possible that you sell us the pheasant?’ (HK'AN_481.1)
b. Bix u k’áat-a’l hun-p’éel tìimbreh
how SBJ.3 ask-PASS.INCMPL one-CL.INAN stamp
ti’ le máak k-u kòon-ol- o’?
LOC DEF person IMPF-SBJ.3 sell\INTRV-INCMPL-D2
‘How does one ask for a stamp from the man who sells [things]?’
(BVS_11.01.22)
In other languages like English or German, action verbs are basically transitive but
possess an optional direct object, i.e. there is no morphological indication of transi-
tivization or detransitivization.
11
In some respects, the central participant exteriorized from intransitive bases by extra-
version is like a cognate object (e.g. with dream, dance) or an object of result, as with
verbs of bodily action/function (e.g. laugh, cough). The respective predicates are said
to be ‘pregnant’ with the participant in question. However, exteriorization of such
closely related internal participants seems to be subject to restrictions in many lan-
guages, e.g. they need to be modified or they do not take regular object marking. This
indicates that these objects do not have the same degree of independence vis-à-vis the
verb as a normal object, including an extraversive object (cf. Lehmann 1991: 192f).
4.3. Applicative vs. extraversive
Extraversion and applicative formation are two processes of undergoer-focused transi-
tivization. They are not in opposition, but rather in complementary distribution on an
asymmetric gradience.
12
The common semantic denominator of both processes may be
paraphrased by ‘direction of action towards an undergoing participant’ (which, inci-
dentally, is also the idea behind the term ‘transitive’). Extraversion is a derivation,
thus a lexical process and not productive for all verbs of some class in a language.
Applicative formation may be productive to different degrees. Some languages have
rather productive applicatives, as is e.g. reported for the Wolof instrument applicative
(Comrie 1985: 318f).
Furthermore, extraversion is restricted to intransitive bases while applicative forma-
tion operates both on intransitive and transitive bases. Thus, extraversion is valence-
11
For English cf. Lemmens 1998, ch. 5.4 on ‘objectless transitives’, who notes that these verbs often
have a restricted set of possible objects which are understood in the intransitive version.
12
The position of the dashed line in Figure 6 is intended to reflect this asymmetry.
Extraversive transitivization in Yucatec Maya
18
increasing, while applicative formation may be valence-increasing (with (in)transitive
base verbs) or valence-rearranging (with transitive base verbs), as established in Com-
rie 1985. While applicative formation promotes adjuncts in diverse relatively periph-
eral roles, extraversion exteriorizes a participant closely related to the situation core
designated by the base verb. It thus serves the individuation of this participant. The
broad concept of applicative is liable to obscure this difference if a participant that
cannot be accommodated in the frame of an intransitive base verb is said to be pro-
moted in the transitivized version, as if ‘no syntactic function’ were the lowest posi-
tion on the hierarchy of syntactic functions Figure 2. This difference between the two
processes is visualized in Figure 5.
Figure 5. Undergoer-focused transitivization
semantic
configuration
situation core
participant
situation core
participant
situation core
participant
semantic
process exteriorization
centralization
construction
intransitive verb
– transitive verb
direct object intransitive verb
adjunct
derivational
process extraversion
appl. formation
The regular relationship between an adjunct construction and an applicative construc-
tion as illustrated, e.g, in (12a) and (12b) is often conceived thus: The dependent in
question is joined to the verb by means of a case relator. In the adjunct construction,
this takes the form of an adposition heading the adjunct. In the applicative construc-
tion, the case relator instead attaches to the verb so that what was the complement of
the adposition now becomes a direct complement of the verb. The same can happen
with different case relators which then correspond to local, benefactive, instrumental
etc. applicatives. Whether or not this can be verified as an historical change in some
languages, it is certainly a useful approach to understand the mechanism of applica-
tive formation. At the same time, this model is not applicable to extraversion because
in the prototypical case, there is no adjunct construction to begin with. Instead, there
is just transfer of an intransitive base into the transitive valence class by means of a
transitivizer.
Christian Lehmann & Elisabeth Verhoeven 19
Finally, since applicative formation is a promotion process, it may play a role in func-
tional sentence perspective, a function that is not accessible to extraversion because
there is no alternative syntactic frame, and thus, no choice of coding. Figure 6 summa-
rizes the differences between the two processes.
Figure 6. Continuum of undergoer-focused valence-increasing operations
lexical syntactic
non-productive productive
base has no direct object slot base may have direct object slot
purely valence-increasing valence-increasing or rearranging
exteriorized roles determined
by lexical base
applied roles determined
by applicative marker
(increased) individuation
of exteriorized object
increased affectedness
of applied object
- feeds topic continuity
extraversive applicative
Some of the parameters mentioned on the left side of Figure 6 hold for both extraver-
sive and applicative formation, while those mentioned on the right side exclusively
characterize applicative formation.
We are now prepared to take a closer look at the thematic roles involved in both proc-
esses. Figure 7 visualizes the semantic space of thematic roles relevant in applicative
and extraversive constructions. The horizontal distribution is arranged according to a
role’s position in the causal chain of an event (following Croft 1991 and Luraghi
2001). Further parameters considered in Figure 7 are the empathy of the participant,
its macrorole and its cross-linguistic frequency in applicative constructions. The em-
pathic roles are positioned in the upper part of Figure 7, the anempathic ones in the
lower part and the neutral ones (theme, patient, stimulus) in between. The macroroles
undergoer and indirectus are included in a box in the same way as the local, the con-
comitant and the antecedent roles.
Extraversive transitivization in Yucatec Maya
20
Figure 7. Focal instances of thematic roles in extraversive and applicative constructions
emitter
addressee
beneficiary
recipient
comitative
theme patient
instrument stimulus
source place goal
Since extraversion is a lexical process, it involves thematic roles that are more or less
inherent in the lexical frame of the base verb. These are prototypically patient and
theme, as indicated by the hatched field in Figure 7. The same goes for the stimulus
with experiential verbs, addressee or theme with communication verbs and place, goal
or source with position and motion verbs.
Applicative formation, on the other hand, is (at least potentially) open to all roles pos-
ited in Figure 7. Following the typological study of Peterson 1999, however, periph-
eral roles such as beneficiary, comitative and instrument are predominant. Applicative
constructions are in general most productive with these latter roles.
13
Furthermore,
these roles may be added to different kinds of situations such as creation or destruc-
tion, action in general, motion, transfer and others. These roles have been included in
the gray area in Figure 7.
5. Assessing Yucatec Maya extraversion
5.1. Frequency
Table 2 shows the distribution of roles over undergoers created by transitivization of
63 YM verbs with the -t suffix. Patient (as in (10)) and theme roles (as in (11))
predominate clearly, followed by the stimulus (as in (4), (8)) and local roles (as in (5),
(9)). Roles subsumed under the indirectus, such as addressee (as in (7)), beneficiary,
13
Cf. again the instrument applicative in Wolof (Comrie 1985).
ANTECEDENT
LOCAL
ANIMATE
INANIMATE
CONCOMITANT
INDIRECTUS
UNDERGOER
Christian Lehmann & Elisabeth Verhoeven 21
recipient, are seldom involved in YM extraversion. Certain thematic roles frequently
involved in applicative formation in other languages, such as the instrument and the
comitative, do not occur at all.
Table 2. Frequency of thematic roles occurring in YM extraversion
14
thematic role frequency
patient 22
theme 17
stimulus 11
goal / source / path
11
place 8
addressee 5
beneficiary 2
recipient 2
5.2. Nature of roles: central vs. peripheral participants
Apart from theme and patient extraversion, other central or closely related participants
may be exteriorized. (22) shows that a verb-inherent local participant such as the path
of a motion may be the direct object of the transitivized verb. Contrast this with (23),
where it is not the local adjunct of (23a) that underlies the direct object of (23b). In
general, peripheral local roles – here a location with respect to a perception verb –
cannot be represented by the direct object of an extraversive verb.
(22) a. h áalkab-nah-en t-u bèel-il in kòol
PRV run-CMPL-ABS.1.SG LOC-POSS.3 way-REL POSS.1.SG milpa
‘I ran on the way to my milpa
15
’
b. t-in wáalkab-t-ah u bèel-il in kòol
PRV-SBJ.1.SG run-TRR-CMPL POSS.3 way-REL POSS.1.SG milpa
‘I ran the way to my milpa’ (AVC_0003/4)
(23) a. le ko’lel-o’ táan u ch’èeneb ti’ le hòol-o’
DEF woman-D2 PROG SBJ.3 peek LOC DEF hole-D2
‘the woman is peeking through the door’ (MPK_031)
b. le ko’lel-o’ táan u ch’èeneb-t-ik
DEF woman-D2 PROG SBJ.3 peek-TRR-INCMPL
(bix yan-il) le hòol-o’
how EXIST:EFOC DEF hole-D2
‘the woman is examining the door’ (MPK_032)
14
Since some verbs allow for more than one participant type as object, the total number of roles is
higher than that of verbs.
15
A milpa is a cleared field, usually located in the rainforest, which is used for agriculture.
Extraversive transitivization in Yucatec Maya
22
Promotion of such peripheral roles is, however, well attested with local adjuncts of ac-
tion verbs in languages with applicative systems, witness (24) from Chichewa
(Mchombo 1998: 506/7) and (25) from Kichaga (Tanzania, Bresnan and Moshi 1993:
49).
16
(24) a. Kalúlú a-ku-phík-á maûngu pa chulu.
1a:hare 1.SBJ-PRS-cook-FV 6:pumpkin 16:on 7:ant.hill
17
‘The hare is cooking some pumpkins on the ant-hill.’
b. Kalúlú a-ku-phík-ír-a pa chulu maûngu.
1a:hare 1.SBJ-PRS-cook-APPL-FV 16:on 7:ant.hill 6:pumpkin
‘The hare is cooking on ant-hill the pumpkins.’
(25) a. N-a
B
-i
B
-ly-à k-élyà.
FOC-1.SG-PRS-eat-FV 7-food
‘He/She is eating food.’
b. N-a
B
-i
B
-lyì-í-à m
a
-rì-nyì k-élyà.
FOC-1.SG-PRS-eat-APPL-FV 3-homestead-LOC 7-food
‘He/She is eating food at the homestead.’
This confirms that YM extraversion chiefly allows for central participants to appear in
direct object function, while applicative formation chiefly allows for peripheral par-
ticipants to appear in direct object function.
5.3. Nature of roles: affected human participants
The direct object of an extraverted intransitive root practically never codes an affected
human participant. The closest one can get is the addressee of a verb of communica-
tion. However, this is not really affected, and even there the construction is marginal.
The canonical frame for a verb of communication is ‘message = direct object, ad-
dressee = indirect object’, as shown in (26a) and (27), first variant. An alternate frame
has the addressee as direct object if there is no NP referring to the theme, as in (26b).
Some speakers allow for absolutive (i.e. direct object) marking of the addressee even
with the basic communication verb tsikbat ‘tell, chat’, as is shown in (27), second
variant, though conservative speakers judge it as ungrammatical. Note that this pattern
is restricted for a number of reasons. First, it violates rules of YM grammar since
there is no cross-reference marking of the lexical NP in object function hun-p’éel ba’l
‘one thing’, the absolutive suffix-slot being occupied by the absolutive marker for the
second person. Furthermore, this frame seems to be conditioned by overt marking of
16
Although the local marking is preserved with the local dependents in the applicative constructions in
(24b) and (25b), these exhibit syntactic properties of a primary object in Bantu, i.e. adjacency to the
verb, passivizability, and possible object cross-reference on the verb (see Alsina and Mchombo 1993,
section 4.3, Bresnan and Moshi 1993).
17
In the Bantu examples, arabic numerals (occasionally followed by a small letter) in front of nouns
indicate noun classes.
Christian Lehmann & Elisabeth Verhoeven 23
the addressee appearing as an absolutive suffix on the verb. Since the third person
singular absolutive suffix is zero in non-clause-final position, a third person (singular)
addressee could not be distinguished from cross-referencing the message argument.
Note furthermore that in any case, the addressee of tsikbat cannot become the subject
of a passive version, from which we may conclude that it is not a canonical direct
object.
18
(26) a. káa in tse’kt te’x u t’àan diyos
CNJ SBJ.1.SG preach:TRR(SUBJ) you.all POSS.3 speech god
‘so that I preach you the word of god’ (ACC)
b. táan in tse’k-t-ik le máak-o’b-a’
PROG SBJ.1.SG preach-TRR-INCMPL DEF person-PL-D1
káa y-ohel-t-o’b
CNJ SBJ.3-know-TRR(SUBJ)-3.PL
ba’x k-u tàal u yúuch-ul
what IMPF-SBJ.3 come SBJ.3 happen-INCMPL
‘I am preaching to these people so that they come to know what is going
to happen’ (AVC_0019)
(27)
táan
in tsikbat-ik tèech hun-
p’éel
ba’l
chat:TRR-INCMPL(ABS.3.SG) you
?
tsikbat-ik-ech
PROG
SBJ.1.SG
chat:TRR-INCMPL-ABS.2.SG
one- CL.INAN thing
‘I am telling you something’ (RMC_1324)
Thus, applicative promotion of the addressee is heavily constrained in several re-
spects: First, in a complete frame displaying message and addressee, the message is
always coded as the direct object while the addressee takes the function of the indirect
object. Second, absolutive coding of the addressee is restricted for basic verbs of
communication, as explained above. It may be of relevance that the most basic verb of
communication, a’l ‘say’, is only used in the canonical frame illustrated in (27), first
variant and excluded from a frame like (27), second variant. We therefore hypothesize
that the addressee in the position of a direct object of an extraversive verb is a secon-
dary development.
5.4. Development: locative alternation
There are a number of contact verbs displaying extraversion together with a valence
alternation. With these either the theme/patient (28a) or the place of contact (28b) is in
18
This is, in fact, the only sentence in our corpus that might lead one to suspect a primary/secondary
object distinction for YM.
Extraversive transitivization in Yucatec Maya
24
direct object function. While there is no alternative intransitive frame for the patient
(28a’), the place may as well be accommodated in an intransitive frame (28b’).
(28) a. táan in ts’íib-t-ik hun-p’éel t’àan màayah
PROG SBJ.1.SG write-TRR-INCMPL one-CL.INAN speech maya
‘I am writing a Mayan word’ (RMC_2011)
a’. *táan in ts’íib ti’ hun-p’éel t’àan
PROG SBJ.1.SG write LOC one-CL.INAN speech
b. táan in ts’íib-t-ik hun-p’éel hu’n / pak’
PROG SBJ.1.SG write-TRR-INCMPL one-CL.INAN paper/brickwork
‘I am writing on a paper / a wall’ (RMC_2011)
b’. táan in ts’íib ti’ hun-p’éel hu’n / pak’
PROG SBJ.1.SG write LOC one-CL.INAN paper / brickwork
‘I am writing on a paper / a wall’ (ACC)
(28) involves two alternations. The alternation between (28b) and (28b’) looks like
applicative formation, as the place argument is promoted to direct object function.
The relationship between (28a) and (28b), however, is a commutation of two partici-
pants in direct object function and thus looks like a semantic valence alternation, one
frame displaying the theme/patient of writing in direct object function, the other dis-
playing the place of writing in direct object function.
(29) suggests that the frame with a local direct object, as in (28b), is secondary or de-
rived. While in (29a) both participants, i.e. the theme and the place, can be accommo-
dated, this is not the case in (29b) (~(28b). Speakers judge the adjunction of the ef-
fected object as infelicitous since the verb ts’íib(t) ‘write’ already implies the ‘object’
t’àan ‘speech, words’ so that its coding in a concomitant phrase seems to be superflu-
ous.
(29) a. táan in ts’íib-t-ik le t’àan ti’ le hu’n-o’
PROG SBJ.1.SG write-TRR-INCMPL DEF speech LOC DEF paper-D2
‘I am writing that word on the paper’ (EMB_0630)
b. táan in ts’íib-t-ik hun-p’éel hu’n
PROG SBJ.1.SG write-TRR-INCMPL one-CL.INAN paper
(
?
yéetel t’àan-o’b)
with speech-PL
‘I am writing on a paper (words)’, lit.: ‘I am writing a paper with words’
(ACC)
Total demotion of the underlying patient may also be observed in genuine applicative
constructions, e.g. in examples like (14), where applicative formation, promoting the
place to direct object function, may be used in order not to mention the patient any-
more. This may be due to the fact that the patient is understood in the context or is
implicit in the verb, so that it would be infelicitous to mention it.
Christian Lehmann & Elisabeth Verhoeven 25
5.5. Development: incorporation plus transitivization
To complete the picture, we have to look at incorporative verbs, i.e. verbs that have an
incorporated noun. These take the transitivizing suffix -t, too.
19
Part of them displays
the rearranging type of applicative construction, in the following form: The underly-
ing direct object is demoted by being incorporated in the verb. The incorporative verb
is intransitive. Subsequent transitivization reopens the direct object slot for another
participant. (30) illustrates promotion of a place adjunct, (31) demonstrates promotion
of a (human) goal adjunct to direct object function.
20
(30) a. t-in ch’ak-ah che’ ichil in kòol
PRV-SBJ.1.SG cut-CMPL tree in POSS.1.SG milpa
b. t-in ch’ak-che’-t-ah in kòol
PRV-SBJ.1.SG cut-tree-TRR-CMPL POSS.1.SG milpa
‘I chopped trees in my cornfield.’ (Bricker, Po’ot Yah and Dzul de Po’ot
1998: 354 / RMC_1685)
(31) a. t-in wek-ah ha’ ti’ h-pìil
PRV-SBJ.1.SG spill-CMPL water LOC M-Philip
b. t-in wek-ha’-t-ah h-pìil
PRV-SBJ.1.SG spill-water-TRR-CMPL M-Philip
‘I threw water on Philip’ (Bricker, Po’ot Yah and Dzul de Po’ot 1998:
354)
In such cases, the function of -t is indistinguishable from the applicative. However,
not all combinations of incorporation plus transitivization are applicative in this sense.
There are a number of lexicalized incorporatives, as in (32a), which do not possess a
regular paraphrase such that the participant underlying their direct object would be
accommodated in their base verb frame. Here, the function is again extraversive, the
intransitive incorporative verb lek’ich being the basis of derivation. The direct object
participant of the transitivized incorporative verb is a new participant arising from the
semantic change brought about by the operation of incorporation.
(32) a. t-u lek’-ich-t-ah u xùun
PRV-SBJ.3 open-eye-TRR-CMPL POSS.3 spouse
b.
?
t-u lek’ah u yich ti’ u xùun
PRV-SBJ.3 open-CMPL POSS.3 eye LOC POSS.3 spouse
‘he looked scornfully at his wife’ (EMB_0176, RMC_2000)
19
For a comprehensive analysis of incorporation and related processes in YM, cf. Lehmann and Ver-
hoeven 2005.
20
Other languages that display rearranging applicative constructions on the basis of incorporation in-
clude Chukchee (cf. Nedjalkov 1976, Spencer 1995), Ainu (cf. Kaiser 1997) and Oluta Popoluca (cf.
Zavala Maldonado 1999).
Extraversive transitivization in Yucatec Maya
26
Summarizing the YM constructions involving derivation in -t, there are only a few
that fulfill the criteria of applicative constructions. Most of these are derived or sec-
ondary with regard to extraversion, or a recent development, as in the case of ad-
dressee-direct-objects.
6. Conclusion
Peterson (1999, section 5.4) postulates the following implicational relationship be-
tween different applicative types: locative and circumstantial applicatives depend on
the presence of other applicative constructions, while benefactive and instrumen-
tal/comitative applicatives do not. That is, the locus of applicative formation is in the
promotion of benefactive and instrumental/comitative arguments. These serve as an-
chors, as it were, for the development of additional applicative constructions marked
either by the same or distinct morphology. This is visualized in the upper half of
Figure 8.
Figure 8. Expansion of constructions with respect to thematic roles
applicative
extraversion
extraversive
The upper half of Figure 8 represents a developmental path for the extension of appli-
cative constructions. The facts of YM now suggest that the picture can be completed
by looking at the process of extraversion with its developmental characteristics. Ex-
traversion and the thematic roles affected by it are shown in the lower half of Figure
8. The locus of extraversion is in the exteriorization of the patient and theme. If other,
more peripheral, roles like goal, place, stimulus etc. join the transitive pattern, this
may take on features of applicative marking, provided there is a regular, transforma-
tional relationship between alternative alignments. Such favorable circumstances may
be provided, e.g., by incorporation or by the locative alternation. Extraversion may
then fulfill functions proper to applicative formation. Thus, a derivational process may
develop from extraversive to applicative passing through the less inherent roles on the
left side of Figure 8. Given that the two processes are related as shown in Figure 6
above, each of them may expand over part of the overall domain of thematic roles.
concomitant
comitative
instrument
affected human
beneficiary
recipient
addressee
local
place, path
goal, source
experiential
stimulus
inherent undergoer
patient
theme
Christian Lehmann & Elisabeth Verhoeven 27
In purely structural terms and disregarding paraphrase relationships, extraversion and
applicative formation are the same thing. We have repeatedly emphasized that the dis-
tinction we are making is gradual. It may be rephrased this way: Undergoer-focused
transitivization is called applicative formation to the extent that it is a syntactic proc-
ess marked on the verb; it is called extraversion to the extent it is a lexical process
with syntactic consequences.
We do not want to be misunderstood. We are not objecting to the attempt to provide a
syntactic analysis of constructions involving processes of word-formation. We have
tried it ourselves. Discovering and formulating regularities is an important facet of the
epistemic interest of linguistics. However, we suspect that in the case of applicative
constructions, transformational relationships have been overemphasized. That is, if
our proposal to distinguish between applicative and extraversive is accepted, then we
expect that many applicatives which figure in linguistic descriptions will turn out to
be more like an extraversive upon critical examination.
21
21
This can be hypothesized on the basis of statements like the following: “The applicative derivation is
seldom fully productive, being normally restricted to a limited set of intransitive roots.” (Dixon and
Aikhenvald 1997: 80). Here features like reduced productivity and restriction to intransitive bases,
which are typical of extraversion, are diagnosed on the applicative instead.
Extraversive transitivization in Yucatec Maya
28
Abbreviations
Morpheme glosses and syntactic categories
1 1
st
person
2 2
nd
person
3 3
rd
person
ABS absolutive
ACT active
AN animate
APPL applicative
ART article
CL classifier
CMPL completive
CNJ conjunction
D1 proximal deixis
D2 distal deixis
DEAG deagentive
DEF definite
EFOC extrafocal
EXIST existential
F feminine
FOC focus
FUT future
FV final vowel
IMP imperative
IMPF imperfective
INAN inanimate
INCMPL incompletive
INT interrogative
INTRV introversive
ITR intransitive
LOC locative
M masculine
PASS passive
PL plural
POSS possessive
PROG progressive
PROHIB prohibitive
PRS present
PRV perfective
PST past
REL relator
SBJ subject
SG singular
SUBJ subjunctive
TRR transitivizer
Text sources
BVS Blair, Robert W. and Vermont-Salas, Refugio 1965-7, Spoken (Yucatec)
Maya. 2 vols. Chicago: University of Chicago, Dept. of Anthropology.
Reprint: Columbia, Miss.: Lucas Brothers, 1979.
HK’AN Dzul Poot, Domingo 1986, “J-k’an yajaw”. In: Cuentos mayas (tomo II).
Edición bilingüe: español – maya. Mérida, Yucatán: Maldonado; INAH,
SEP; 89-114.
HNAZ Bolio, Antonio 1930, “H Nazario”. Andrade, Manuel J. and Máas Collí,
Hilaria (eds) 1991, Cuentos mayas yucatecos; Tomo II, Mérida: Universidad
Autónoma de Yucatán; 64-127.
ACC Amedée Colli Colli, Felipe Carrillo Puerto, Q. Roo, Mexico,
AME Antonio May Ek, Yaxley, Q. Roo, Mexico,
AVC Aniceto Velázquez Chi, F. Carrillo Puerto, Q., Mexico,
CPP Catalino Poot Peña, Yaxley, Q. Roo, Mexico
EMB Ernesto May Balam, Yaxley, Q. Roo, Mexico
Christian Lehmann & Elisabeth Verhoeven 29
MPK María Puk Ciau, Yaxley, Q. Roo, Mexico
NMP Norma May Pool, Yaxley, Q. Roo, Mexico
RMC Ramón May Cupul, Yaxley, Q. Roo, Mexico
References
Alsina, A. and Mchombo, S.A. 1993. “Object asymmetries and the Chichewa applica-
tive construction”. In Theoretical aspects of Bantu grammar, S. Mchombo (ed.),
17-45. Stanford: Center for the Study of Language and Information [CSLI lecture
notes 38].
Baker, M. 1988. Incorporation. A theory of grammatical function changing. Chicago
etc.: University of Chicago.
Bricker, V., Po’ot Yah, E. and Dzul de Po’ot, O. 1998. A dictionary of the Maya lan-
guage as spoken in Hocabá, Yucatán. Utah: University of Utah Press.
Bohnemeyer, J. 2002a. The grammar of time reference in Yucatec Maya. München:
L
INCOM
Europa [L
INCOM
Studies in Native American Linguistics 44].
Bohnemeyer, J. 2002b. Activity nouns, unaccusativity, and argument marking in Yu-
katekan. Talk held at the SSILA meeting; LSA; San Francisco; 2002.
Bohnemeyer, J. 2004. “Split intransitivity, linking, and lexical representation: The
case of Yukatek Maya”. Linguistics 42: 67-107.
Bresnan, J. and Moshi, L. 1993. “Object asymmetries in comparative Bantu syntax”.
In Theoretical aspects of Bantu grammar, S. Mchombo (ed.), 47-91. Stanford:
Center for the Study of Language and Information [CSLI lecture notes 38].
Comrie, B. 1985. “Causative verb formation and other verb-deriving morphology”. In
Language typology and syntactic description, Vol.III: Grammatical categories
and the lexicon. (repr. 1990), T. Shopen (ed.), 309-348. Cambridge etc.:
Cambridge University Press.
Croft, W. 1991. Syntactic categories and grammatical relations. Chicago: Chicago
University Press.
Dixon, R.M.W. and Aikhenvald, A.Y. 1997. “A typology of argument-determined
constructions”. In Essays on language function and language type. Dedicated to
T. Givón, J. Bybee, J. Haiman and S.A. Thompson (eds), 71-113. Amsterdam and
Philadelphia: John Benjamins.
Dixon, R.M.W. and Aikhenvald, A.Y. 2000. “Introduction”. In Changing valency:
Case studies in transitivity, R.M.W. Dixon and A.Y. Aikhenvald (eds), 1-29.
Cambridge: Cambridge University Press.
Drossard, W. 1991. “Transitivität (vs. Intransitivität) und Transitivierung (vs. Intransi-
tivierung) unter typologischem Aspekt”. In Partizipation: Das sprachliche Erfas-
sen von Sachverhalten, H. Seiler and W. Premper (eds), 408-445. Tübingen: Narr
[Language Universal Series 6].
Extraversive transitivization in Yucatec Maya
30
Givón, T. 1983. “Topic continuity in discourse: An introduction”. In Topic continuity
in discourse: a quantitative cross-language study, T. Givón (ed.), 1-41. Amster-
dam and Philadelphia: John Benjamins.
Foley, W.A. and Van Valin, R.D. 1984. Functional syntax and universal grammar.
Cambridge: Cambridge University Press [Cambridge Studies in Linguistics 38].
Haspelmath, M. 1993. “More on the typology of inchoative/causative verb alterna-
tions”. In Causatives and transitivity, B. Comrie and M. Polinsky (eds), 87-120.
Amsterdam and Philadelphia: John Benjamins [Studies in Language Companion
Series 23].
Haspelmath, M. and Müller-Bardey, T. 2004. “Valence change”. In Morphology. Vol.
2, G. Booij et al. (eds), 1130-1145. Berlin and New York: de Gruyter.
Hopper, P. and Thompson, S. 1980. “Transitivity in grammar and discourse.” Lan-
guage 56: 251-299.
Kaiser, L. 1997. “The interaction of noun incorporation and applicative formation in
Ainu”. In Yearbook of Morphology 1997, G. Booij and J. van Marle (eds), 157-
178. Boston: Kluwer Academic Publishers.
Kaufman, T. 1990. “Algunos rasgos estructurales de los idiomas Mayances”. In Lec-
turas sobre la lingüística Maya, N.C. England and S.R. Elliot (eds), 59-114. La
Antigua, Guatemala: Centro de Investigaciones Regionales de Mesoamérica.
Lehmann, Ch. 1991. “Predicate classes and P
ARTICIPATION
”. In Partizipation: Das
sprachliche Erfassen von Sachverhalten, H. Seiler and W. Premper (eds), 183-
239. Tübingen: Narr [Language Universal Series 6].
Lehmann, Ch. 1993. “Predicate classes in Yucatec Maya”. Función (Guadalajara)
13/14: 195-272.
Lehmann, Ch. 2002. Possession in Yucatec Maya. Erfurt: Seminar für Sprachwissen-
schaft der Universität [ASSidUE 10]. Second revised edition.
Lehmann, Ch., Shin, Y.-M. and Verhoeven, E. 2000. Direkte und indirekte Partizipa-
tion: Zur Typologie der sprachlichen Repräsentation konzeptueller Relationen.
München: L
INCOM
Europa [L
INCOM
Studies in Language Typology 4].
Lehmann, Ch. and Verhoeven, E. 2005. “Noun incorporation and participation. A
typological study on participant association with particular reference to Yucatec
Maya”. In Aspects of participation: Typological studies on concomitance and in-
corporation, Ch. Lehmann (ed.), 103-204. Berlin: Akademie Verlag [Studia
Typologica 7].
Lemmens, M. 1998. Lexical perspectives on transitivity and ergativity: causative con-
structions in English. Amsterdam and Philadelphia: John Benjamins [Current is-
sues in linguistic theory 166].
Luraghi, S. 2001. “Syncretism and the classification of semantic roles”. STUF 54: 35-
51.
Mchombo, S.A. 1998. “Chichewa (Bantu)”. In The handbook of morphology, A.
Spencer and A.M. Zwicky (eds), 500-520. Oxford and Massachusetts: Blackwell.
Christian Lehmann & Elisabeth Verhoeven 31
Michaelis, L. and Ruppenhofer, J. 2001. Beyond alternations: A constructional model
of the German applicative pattern. Stanford: CSLI Publications.
Mosel, U. 1991. “Towards a typology of valence”. In Partizipation: Das sprachliche
Erfassen von Sachverhalten, H. Seiler and W. Premper (eds), 240-251. Tübingen:
Narr [Language Universal Series 6].
Nedjalkov, V.P. 1976. “Diathesen und Satzstruktur im Tschuktschischen”. In Satz-
struktur und Genus verbi, R. Lötzsch and R. RĤžiþka (eds), 181-211. Berlin: A-
kademie-Verlag [Studia Grammatica XIII].
Nichols, J., Peterson, D.A. and Barnes, J. 2004, “Causativizing and decausativizing
languages”. Linguistic Typology 8: 149-211.
Owen, M.G. 1973. “Semantic aspects of Yucatec dual-object constructions”. In Mean-
ing in Mayan language. Ethnolinguistic studies, Edmonson (ed.), 51-57. The
Hague and Paris: Mouton [Jan. ling. Ser. Pract 158].
Paris, C. 1985. “Relations actancielles et valence verbale en avar: effacement de
l’actant et mise au jour du sens”. Actances 1: 135-153.
Peterson, D.A. 1999. Discourse-functional, historical, and typological aspects of ap-
plicative constructions. Dissertation, Berkeley: University of California.
Rice, S.A. 1987. Towards a cognitive model of transitivity. Dissertation, San Diego:
University of California.
Rude, N. 1986. “Topicality, transitivity, and the direct object in Nez Perce”. Interna-
tional Journal of American Linguistics 52: 124-153.
Shibatani, M. 1996. “Applicatives and benefactives: A cognitive account”. In Gram-
matical constructions: Their form and meaning, M. Shibatani and S.A. Thompson
(eds), 157-194. Oxford: Clarendon Press.
Spencer, A. 1995. “Incorporation in Chukchi”. Language 71(III): 439-489.
Tsunoda, T. 1981. “Split case-marking patterns in verb types and tense/aspect/mood”.
Linguistics 19: 389-438.
Van Valin, R.D. and LaPolla, R.J. 1997. Syntax: Structure, meaning and function.
Cambridge etc.: Cambridge University Press.
Zavala Maldonado, R. 1999. “External possessor in Oluta Popoluca (Mixean). Appli-
catives and incorporation of relational terms”. In External possession, D. Payne
and I. Barshi (eds), 339-372. Amsterdam and Philadelphia: John Benjamins.
