Journal Code
L N C 3
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
Article ID
Dispatch: 06.03.14
1 2 0 7 9 No. of Pages: 17
CE: Sapon, Abbey Valerie
ME:
Language and Linguistics Compass (2014): 1–17, 10.1111/lnc3.12079
Direct/Inverse Systems
Q1
Guillaume Jacques* and Anton Antonov
CNRS-INALCO-EHESS, CRLAO
1. Introduction
The term ‘inverse’ has been used in the typological literature to designate a considerable
variety of phenomena, which have been analyzed in several distinct frameworks. Direct–
inverse systems are attested in a number of endangered languages spoken in the Americas
(Algonquian, Mapudungun, Sahaptian, etc.) and the Himalayas (Sino-Tibetan).
The aim of this article is to provide an overview of direct/inverse systems in a perspective
as free as possible from framework-specific assumptions using the most recent data and
adopting both a typological and a historical perspective. It is divided into four sections.
First, we propose a definition of the canonical direct/inverse system and introduce the
concepts of proximate/obviative and referential hierarchies.
Second, we present examples of attested direct/inverse systems in the world’s languages by
subdividing them into two main categories: near-canonical and highly non-canonical, and
additionally discussing systems with hierarchical agreement but without direction (direct/
inverse) marking.
Third, we evaluate to what extent direct/inverse systems are correlated with other typological features.
Fourth, we apply a panchronic perspective on direct/inverse systems, studying their
attested origin and their evolution and how the diachronic pathways can help understand
the present data.
2. Basic Definition
Before defining the canonical direct/inverse system, we need to introduce some of the terminology used by typologists to describe systems which index two arguments.1 It is customary to represent these systems using semantic maps in tabular format as in Table 1, where T1
rows indicate agent and columns patient. The different transitive configurations are symbolically represented by using an arrow, with the agent on its left and the patient on its right,
both abbreviated as 1, 2, 3 for first, second, and third person, respectively. The first and
second persons are called SPEECH ACT PARTICIPANTS (SAP). In the case of third person
arguments, 3 indicates proximate and 3′ obviative referents (2.2). In intransitive forms, by contrast,
the abbreviation refers to the sole argument of the verb. They are systematically included for
reference.
The cells corresponding to the 1→1 and 2→2 configurations are semantically reflexive
and are thus filled in gray, since in most languages, they tend to be expressed by an intransitive construction.2 The 3→3 cell, on the other hand, is not since the corresponding configuration is not necessarily reflexive.
This table does not represent number, clusivity, and obviation, but more complex paradigms including these features will be studied below.
As will become clear in the course of this article, it is convenient to separate the transitive
paradigm into three DOMAINS (Zúñiga 2006, 47-54), represented in Table 1 by different
colors. First, the LOCAL domain (in blue) comprises the forms 1→2 and 2→1, where both
© 2014 The Authors
Language and Linguistics Compass © 2014 John Wiley & Sons Ltd
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
2
Guillaume Jacques and Anton Antonov
Table 1. The three domains of the transitive paradigm.
arguments are SAPs. Second, the NON-LOCAL domain (in red) refers to the cases where both
arguments are third person. Third, the MIXED domain (in green) includes all the forms with a
SAP argument and a third person (1→3, 2→3, 3→1, 3→2).
2.1. THE CANONICAL DIRECT/INVERSE SYSTEM
Given the diversity of direct/inverse systems in the world’s languages, canonical typology
(Corbett 2007) seems to be a useful tool for accurately describing them by accounting for
their deviation from the canon. The canonical direct/inverse system could then be defined
as a type of transitive person marking system presenting three essential characteristics.
First, in such a system, all person–number markers are neutral with regard to syntactic roles
(S, A, and O).3
Second, the ambiguity which this entails (especially in mixed scenarios) is resolved by way
of obligatory (and mutually exclusive) markers, called direct (in the case of 1→2, SAP→3,
3→3′) and inverse (in the case of 2→1, 3→SAP, and 3′→3), respectively. These markers
do not appear on intransitive verbs. This property is generally described in terms of
referential hierarchies (see Section 2.3).
Third, inverse verb forms, unlike verb forms in a passive construction, do not undergo
valency changes (the verb does not become intransitive when an inverse marker is added),
and the arguments keep the same syntactic properties (such as case marking and pivot
accessibility) as in the direct construction.
Table 2 illustrates the canonical system, for which we have taken 1 > 2 > 3 > 3′ as the T2
canonical person hierarchy. We can observe the perfectly symmetrical distribution of direct
and inverse forms, along an axis running from the upper left side to the bottom right side
of the table. Direct forms appear in orange and inverse ones in green.
No language exactly attests the pattern in Table 2, although Rgyalrong languages
(Sino-Tibetan, Sichuan, China) are among the closest to it.4
Table 2. The canonical direct/inverse system.
© 2014 The Authors
Language and Linguistics Compass © 2014 John Wiley & Sons Ltd
Language and Linguistics Compass (2014): 1–17, 10.1111/lnc3.12079
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
Direct/Inverse Systems
3
The status of the local and non-local domains varies considerably across languages, but in
most languages with near-canonical systems, inverse and/or direct markers can also be found
in these domains.
In the local domain, all possibilities but one appear to be attested: 2→1 receives the inverse
marker (as in Situ Rgyalrong), both 2→1 and 1→2 receive the inverse marker (as in
Mapudungun, a language isolate of Chile), or else 1→2 and 2→1 are expressed by special
forms unrelated to either the direct or the inverse ones (as in Algonquian). Another possibility is that the use of direct/inverse markers in the local domain can depend on the particular
person–number combination involved (as in Kiranti languages). The one possibility not
attested to the best of our knowledge is one where 1→2 would receive inverse marking,
whereas 2→1 would be marked as direct.
In the non-local domain, some languages with a direct/inverse system present a further
contrast between proximate and obviative referents (see Section 2.2 for more details). In such
cases, the inverse marker appears in the 3′→3 and the direct marker in the 3→3′ configuration, and the same direction markers are shared between the mixed and non-local domains.
2.2. PROXIMATE/OBVIATIVE
Algonquian languages have grammaticalized a distinction between two types of third person
referents. These are generally designated as proximate (3) and obviative (3′). This distinction is
redundantly marked on both nouns and all verb forms, including transitive and intransitive
ones (see Section 3.1.2). In addition, the proximate/obviative status of the arguments determines the presence of direct or inverse marking on transitive verbs: in non-local scenarios,
when the A is proximate and the O is obviative, the direct form is used, whereas in the case
when the A is obviative and the O is proximate, the inverse form is required. In example 1,
the verb is in the direct form, as the A is proximate and hence unmarked, while the O is
obviative (marked with the -a suffix), whereas in example 2, the converse is true.
many kill-3→3′ badger buffalo-OBV
‘Badger (3) has killed many buffalo (3′)’.
scare-INV man dog-OBV
The dog scares the man. (Wolfart 1973, 25)
In the mixed domain, the proximate/obviative status of the third person referent is also
indicated on the verb (see Table 3).
T3
No language outside of Algonquian displays a system in which the proximate/obviative
distinction is marked on nouns, on intransitive verbs, and in the mixed domain. Yet, for
the purpose of typological comparison, we find it legitimate to use the terms ‘proximate’
and ‘obviative’ to describe systems which distinguish two third person referents in the
non-local domain only by means of direct/inverse marking on the verb form.
Indeed, in most languages with direct/inverse systems, unlike Algonquian, the presence of
direct or inverse markers on the verb is the only clue to the proximate or obviative status of a
particular argument in non-local scenarios. In example 3, from Japhug Rgyalrong for instance,
the absence of inverse marker on the verb indicates a direct form, which implies that the A
(the maid) is proximate, while the O (the rooster) is obviative. Inverse marking here would indicate the opposite scenario (A obviative and O proximate) without changing the syntactic roles.
© 2014 The Authors
Language and Linguistics Compass © 2014 John Wiley & Sons Ltd
Language and Linguistics Compass (2014): 1–17, 10.1111/lnc3.12079
1
4 Guillaume Jacques and Anton Antonov
2
3 Q3 Table 3. Plains Cree present paradigms. TA wâpam- ‘see’ and IA pimipahtâ- ‘run’ (Wolfart 1973).
4
5
6
7
8
9
10
11
12
13
14
15
16
servant TOP ERG rooster TOP EVD-kill
17
‘The
maid killed the Rooster’. (Japhug, Aesop adaptation ‘The old woman and the maid’)
18
19
As is the case in Japhug, in languages with direct-inverse contrast in the non-local domain,
20
the choice of proximate or obviative status for a particular referent (and thus the use of an
21
inverse or a direct form) is guided by both semantics (relative animacy of the agent and patient)
22
and pragmatics (the relative topicality or saliency of the two referents).
23
24
2.3. REFERENTIAL HIERARCHIES
25
26
Some authors have proposed to apply the (pre-existing) concept of referential hierarchies in order
27
to provide a unified account of the use of direct/inverse markers in all three domains, including
28
the question of how the proximate or obviative status of the referents is determined.
29
Since DeLancey (1981a, 644), direct/inverse systems are often described in terms of the
30
Empathy Hierarchy presented in (4)5:
31
(4) SAP > third person pronoun > human > animate > natural forces > inanimate
32
33
The distribution of direct and inverse markers can then be captured by the following rule:
34
(5) If the patient is higher on the hierarchy than the agent, the verb receives inverse marking; con35
versely, if the agent is higher on the hierarchy than the patient (or if both are equal), the verb
36
receives direct marking.
37
38
In DeLancey’s view, the Empathy Hierarchy in (4) is motivated by several factors.
39
First, the ranking SAP→3 is due to an ‘inherent natural attention flow’, whereby discourse
40
participants (first or second person) are the ‘natural starting point’ of most utterances. Second,
41
the ranking within the third persons is explained by the concept of empathy, hence the name
42
of the proposed hierarchy: ‘speakers, being animate and humans, are more likely to empa43
thize with (i.e. take the viewpoint of) human beings than animals, and of animals than
44
inanimates’.
45
The hierarchy in (4) is a general model which has to be adapted to each language to cor46
rectly predict the attested forms, and some variation can be observed even between closely
47
related languages.
48
A detailed investigation of hierarchies in direct/inverse systems is beyond the scope of this
49
paper (see Zúñiga 2006 for more detail). Overall, the two rankings SAP > 3 and animate
50
inanimate appear to be relatively robust and can even be considered definitional of direct/
51
inverse systems within a theoretical framework using referential hierarchies.6
52
53
© 2014 The Authors
Language and Linguistics Compass (2014): 1–17, 10.1111/lnc3.12079
Language and Linguistics Compass © 2014 John Wiley & Sons Ltd
54
55
56
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
Direct/Inverse Systems
5
Additionally, direct/inverse systems are considered by some authors as a special case of
hierarchical alignment (Nichols 1992), a type distinct from accusative, ergative, active, or tripartite alignment. Following Siewierska (1998, 10), hierarchical alignment can be defined as an
alignment type where ‘the treatment of the A and O is dependent on their relative ranking
on the referential and/or ontological hierarchies’.
3. Attested Systems
This section will present three types of hierarchical systems: near-canonical direct/inverse
systems, highly non-canonical direct/inverse systems, and hierarchical systems without direction (direct/inverse) markers.
3.1. NEAR-CANONICAL DIRECT/INVERSE SYSTEMS
3.1.1. Rgyalrong Languages
The direct/inverse systems of Rgyalrong languages have been described in several studies
(DeLancey 1981b on Situ, Sun and Shidanluo 2002 on Tshobdun, Jacques 2010 on Japhug,
and Gong on Zbu). These languages index not only the person but also the number of the
two arguments in the verbal morphology, and the resulting paradigm is more complex than
the pattern shown in Table 2.
In 6, the verb has a direct form, and the inverse is blocked, as the A is SAP, while the O is
third person. On the other hand, 7 illustrates the compulsory use of the inverse marker when
the agent is third person and the patient SAP.
I ERG that AOR-hit:1SG
I hit him/her.
yesterday 3SG one-time you AOR-2-INV-hit
(S)he hit you once yesterday.
The presence or absence of the inverse prefix in 3→3 forms (in other words, the 3′→3 vs
3→3′ contrast) can be predicted by the following rules (Sun and Shidanluo 2002; Jacques 2010):
1. When the agent is animate and the patient inanimate, the inverse is blocked.
2. When the agent is inanimate and the patient animate, the inverse is required.
There is a strong, though not absolute, tendency for the inverse to appear when the agent
is non-human animate and the patient human. When both arguments are human, there is no
absolute constraint blocking or requiring the use of the inverse. 3′→3 forms with inverse
marking are relatively rare in comparison with direct 3→3′ ones and are used when the
patient is markedly more topical than the agent.
Table 4 presents the non-past paradigm of a dialect of Zbu Rgyalrong. The symbols Σ1 T4
and Σ3 represent different stem forms (stem 1 and stem 3, respectively). Stem 3 only occurs
in a subset of direct forms, i. e. those where the patient is third person and the agent is
singular ([123]SG→3(′)) and could thus be considered as a non-canonical direct marker.
In this paradigm, considering only the mixed and non-local forms and leaving aside stem
alternations (Σ1 vs Σ3), there is perfect symmetry between direct and inverse forms, which are
distinguished only by the presence or absence of the inverse prefix we-. Mixed and non-local
direct forms are identical to the corresponding intransitive forms except for the stem
© 2014 The Authors
Language and Linguistics Compass © 2014 John Wiley & Sons Ltd
Language and Linguistics Compass (2014): 1–17, 10.1111/lnc3.12079
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
6
Guillaume Jacques and Anton Antonov
Table 4. Zbu Rgyalrong transitive paradigm (data adapted from Gong ).
alternations. Only 1→2 forms are distinct from all the rest, with a synchronically opaque
portmanteau t - 1→2 prefix (if the system was perfectly symmetrical, 1→2 forms such as
*tə-Σ1-ŋ would be expected).
To account for the Zbu Rgyalrong data presented above, the following hierarchy has to
be posited:
(8) 1 > 2 > 3 animate proximate > 3 animate obviative > 3 inanimate
Applying rule 5 to the hierarchy in (8) predicts the occurrence of the inverse in 2→1,
3→SAP, and 3′→3 forms and its absence in 1→2, SAP→3, and 3→3′. Here, proximate/
obviative (or topical/non-topical) is more relevant than human/non-human as listed in the
original formulation of the hierarchy in (4).
Describing Japhug Rgyalrong data in this framework would require a different hierarchy
(Jacques 2012b):
(9) SAP > 3 animate proximate > 3 animate obviative > 3 inanimate > generic human
In Japhug, there is no inverse in 2→1 forms, while it appears with a generic human agent
(disregarding the animacy of the patient).
Zbu Rgyalrong also allows us to illustrate the ways in which inverse marking can be
shown to differ from passive marking. Indeed, the Zbu prefix wə- cannot be analyzed as a
passive marker for several reasons.
First, it is obvious that the verb still remains transitive even in inverse forms, since especially in 2/3→1SG the number markers -nd ə (dual) or - ə (plural) are suffixed to the first
person -ŋ, and since in all 2→1 forms, both the second person tə- prefix and the first person
-ŋ, -t ə, or -jə suffixes are present. Given the fact that both the agent and the patient are
indexed, one can deduce that the inverse wə- has no intransitivizing effect.
Second, the addition of the inverse prefix has no effect on case marking: third person
agents still receive ergative case, as illustrated by example (10) from Gong ().
Bkrashis stone ERG AOR-INV-hit NON.VISUAL
‘A stone hit bKrashis’. (The stone falls from the mountain, for example)
Third, the inverse prefix wə- is not a derivation contrary to the passive construction in this
language: the semantics of inverse forms are always predictable from that of the base forms,
unlike derivation prefixes in Rgyalrong languages (including passive, anticausative,
antipassive etc; see Jacques 2012b) which present many irregularities.
Few direct/inverse systems are as symmetrical as the one attested in Zbu Rgyalrong, and in
fact, these systems present tremendous diversity across languages. Partially, opaque inverse
systems will be studied in more detail in Section 3.2.
© 2014 The Authors
Language and Linguistics Compass © 2014 John Wiley & Sons Ltd
Language and Linguistics Compass (2014): 1–17, 10.1111/lnc3.12079
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
Direct/Inverse Systems
7
3.1.2. Algonquian Languages
Algonquian languages differ from Rgyalrong and all other languages with direct/inverse
(apart from Kutenai) in that the proximate/obviative distinction is largely independent of
direction marking on the verb.
In Algonquian languages, all animate nouns are marked for proximate and obviative, and
all animate third person referents have a clear status as proximate or obviative in a given sentence. No such distinction is found in the case of SAP (SAPs are never obviative) nor in the
case of inanimate nouns in most of these languages. Furthermore, the singular/plural distinction is neutralized in obviative forms (except in some Central Algonquian languages such as
Fox and Miami-Illinois) or else there is no contrast between proximate and obviative in the
plural (as in Blackfoot). In a single sentence, at most one argument can be proximate, but it is
also possible to find sentences with only obviative arguments.
Table 3 (adapted from Wolfart 1973) illustrates an example of a transitive animate paradigm (the verb wâpam- ‘see’) and an intransitive animate one with pimipahtâ- ‘run’. The direct
markersare -â- or -ê- and the inverse marker is -ik(o). Please note that the proximate/obviative
distinction is also marked on intransitive verbs.
On transitive verbs, the proximate/obviative appears in the mixed domain: there are
different forms for SAP ↔ 3 PROX and SAP ↔ 3 OBV. In the non-local domain, three configurations with direct and inverse marking are observed: 3→3′, 3′→3, and 3′→3′. While
the first two are attested in other languages with direct/inverse systems, the third one is only
found in Algonquian. Here, 3′→3′ (as in the form wâpam-êyiwa ‘heobv sees himobv’) is the socalled further obviative form, which is often abbreviated as 3′ → 3″. Example 11 illustrates this
form with the verb nipah- ‘to kill’.
one just kill-3′→3′ 3POSS-oldersister-OBV rabbit-OBV
‘His sister had killed but one rabbit’.
Such a form illustrates the fact that a given sentence can have any number of obviative arguments. Thus, a configuration with two obviative arguments does not imply reflexivization.
The morphosyntactic marking of the proximate/obviative distinction presents some noteworthy differences with other languages with direct/inverse such as Rgyalrong.
First, in Algonquian, we observe a tripartite distinction between proximate animate, obviative
animate, and inanimate referents. Indeed, the empathy hierarchy for Cree would be the following7:
(12) SAP > animate proximate > animate obviative > inanimate
Second, the proximate/obviative distinction in Rgyalrong is only present in 3→3′ and
3′→3 verb forms, while in Algonquian, it is also redundantly marked on nouns, intransitive
verbs, and transitive mixed forms.
Third, in Algonquian languages, the possessee of a SAP possessor can be either proximate or
obviative, but the possessee of a third person proximate referent must be obviative. Thus, in example 13, where the possessee (‘John’s dog’) is the agent and the possessor (‘John’) is the patient,
the inverse suffix -ik is required, since the former (‘John’s dog’) is automatically obviative.
John 3SG.POSS-dog-OBV PST-bite-INV
Johnprox’s dogobv bit himprox. (Wolfart 1973, 25)
Although a similar constraint has been described in other language families such as Mayan
(see Aissen 1997), no such phenomenon is attested in other languages with direct/inverse in
© 2014 The Authors
Language and Linguistics Compass © 2014 John Wiley & Sons Ltd
Language and Linguistics Compass (2014): 1–17, 10.1111/lnc3.12079
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
8
Guillaume Jacques and Anton Antonov
non-local forms, such as Rgyalrong languages, for instance (see Jacques 2010, 141-2), and is
thus not a universal property of proximate/obviative marking.
This comparison of Rgyalrong and Algonquian shows that while the proximate/obviative
distinction may appear to be a subset of the direct/inverse in non-local forms, the two phenomena are fundamentally different and only intersect in the 3→3′ and 3′→3 configurations. Nearly all languages whose direct/inverse system includes the non-local domain
behave like Rgyalrong languages. The only exceptions we are aware of are Algonquian
and Kutenai, but even the latter has a much simpler system than Algonquian. This also confirms the fact that the Algonquian direct/inverse system, while undeniably the first to have
been properly described, is by no means representative.
3.2. HIGHLY NON-CANONICAL DIRECT/INVERSE SYSTEMS
Highly non-canonical inverse systems comprise two main types: systems with the direct/inverse
contrast present in the local or non-local domains but not in the mixed domain or systems
where the direct or inverse markers are polyfunctional. For lack of space, we will only illustrate
the second case, using data from Khaling, a Kiranti language of Nepal (Jacques et al. 2012).8
As can be seen in Table 5, Khaling has a prefix ?i- which appears in the second person T5
forms of intransitive verbs as well as in the following transitive configurations: 2→1, 3→1,
3→2, and 2→3. Additionally, it appears in 1DU/PL→2 forms which are not shown in that
table. Khaling has no contrast between 3 and 3′.
The prefix ?i- can be described as appearing in all forms involving the second person
(except 1SG→2) and in all inverse forms. It is thus a conflation of a second person marker
and an inverse marker, which has little synchronic functional motivation and cannot be
described in terms of a hierarchy, even though a historical explanation seems possible (see
Section 2). We could gloss the Khaling prefix as 2/INV, but it would be misleading to refer
to it as an inverse marker.
Prefixes with a distribution nearly identical to that of Khaling ?i- are found in related
languages such as Dumi, Trung, and Rawang (van Driem 1993).
3.3. HIERARCHICAL AGREEMENT WITHOUT DIRECTION MARKING
Hierarchical systems without direct/inverse marking can be illustrated with data from Icari
Dargwa (Nakh-Daghestanian, Sumbatova and Mutalov 2003).
This language illustrates the fact that the access to inflectional slots may be described in
terms of an indexability hierarchy. There are two clitics (=di ‘2s’ and = da ‘1, 2p’) which mark
person in several TAM forms, such as the preterite. Second person arguments are always
indexed on the verb irrespective of their grammatical role (S, A, or O). On the other hand,
1st person arguments are indexed only if there is no 2nd person involved (see 14 vs 15)
Table 5. Khaling verbal system (singular forms only).
© 2014 The Authors
Language and Linguistics Compass © 2014 John Wiley & Sons Ltd
Language and Linguistics Compass (2014): 1–17, 10.1111/lnc3.12079
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
Direct/Inverse Systems
9
1SG-ERG 2SG(ABS) catch.M.PFV-PRET=2SG
I caught you.
1SG-ERG M. catch.M.PFV-PRET=1SG
I caught Murad.
Finally, when there are no SAP arguments, none of these clitics is used. In other words,
person indexability is only determined by the hierarchy 2 > 1 > 3 not by grammatical roles
(see Table 6 with the verb uc- ‘to catch’ in the preterit form -ib).
T6
The paradigm in Table 6 is the prime example of a hierarchical system without a direction
marker.
However, this simple picture looks different when TAM categories derived from the
thematic stem are taken into account (Sumbatova and Mutalov 2003, 81-82, 99).
The choice of the thematic vowel follows the hierarchy SAP > non-SAP. Thus, -i- shows
that the AGENT is higher on the hierarchy than the PATIENT, -u- shows that this is not the
case, i.e. -i- is used if the AGENT is SAP and the patient isn’t, and -u- in all other cases
(including intransitive forms, except with 3 person subjects where -ar-/-an- is used instead).
Person markers in the thematic TAM categories cross-reference only SAP: we have -d ‘1st
person (SG/PL)’ and -t/-ta ‘2nd person (SG/PL)’, the 3rd person being unmarked (cf. Table 7). T7
Although pure hierarchical systems without any direct or inverse marker are well attested,9
it is possible that many languages with hierarchical alignment appearing to lack direction
markers may reveal restricted direct or inverse marking in some areas of their grammar, like
Icari Dargwa. In particular, Tangut (Sino-Tibetan; extinct) is generally presented as having a
verbal agreement system very close to that of Dargi in Table 6 (the only significant difference
being that in Tangut in local forms the verb agrees with the patient, thus 2→1 receives the
first person marker instead of the second person marker as in Dargi, cf Kepping 1981). However, it has more recently been shown that Tangut consistently distinguishes SAP.SG→3 and
3→SAP.SG by stem alternation in the direct forms (Jacques 2009).
Table 6. Icari Dargwa person markers in the preterite.
1
1
2
3
uc-ib = di
uc-ib = da
2
3
uc-ib = di
uc-ib = da
uc-ib = di
uc-ib
uc-ib = di
Table 7. Icari Dargwa potential present verb paradigm.
AP
1
1
2s
2p
3
urc-u-t
urc-u-t-a
urc-u-d
© 2014 The Authors
Language and Linguistics Compass © 2014 John Wiley & Sons Ltd
2s
2p
3
urc-u-t
urc-u-t-a
urc-u-t
urc-u-t-a
urc-i-d
urc-i-t
urc-i-t-a
urc-u
Language and Linguistics Compass (2014): 1–17, 10.1111/lnc3.12079
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
10 Guillaume Jacques and Anton Antonov
4. Typological Perspectives
An important question concerning direct/inverse systems is to what extent this typological
feature is correlated with other features. Klaiman (1991, 1992, 165-169) has proposed a
correlation with head-markedness and configurationality, which was indeed observed in
most of the languages in her sample.
More than twenty years after Klaiman’s work, it seems unlikely that direct/inverse could
be correlated with any other feature (or set of features) in particular. Table 8 presents data T8
from Zbu (Gong ), Cree (Wolfart), Sahaptin (Rude 1997; Zúñiga 2006), Navajo (Willie
2000), Dargwa (Sumbatova and Mutalov 2003), Movima (Haude 2009), and Mapudungun
(Zúñiga 2006).
This table indicates whether the inverse in these languages appears in the local, non-local,
or mixed domains, as well as the presence or absence of other typologically relevant features
such as head-markedness, discontinuous constituents, case marking, polypersonal marking on
the verb, incorporation, and strong noun–verb distinction.
These data show that languages with direct/inverse systems, though mainly head-marking
and polysynthetic (i.e. having noun incorporation and polypersonal marking), need not be so.
It is important to note that non-configurationality, if understood as allowing discontinuous
constituents, is likewise rare among such languages.
Given the limited number of languages with direct/inverse systems, it may prove difficult
to ascertain whether the presence of such a system in a given language could be correlated
with the presence or absence of other features. Indeed, even if we surveyed all such
languages, the resulting sample would still be too limited to provide any statistically significant
results, especially if all languages of a particular subgroup (Algonquian or Rgyalrong, for instance)
are counted as one token.
4.1. EXTENSION OF THE CONCEPT OF DIRECT/INVERSE
As defined in Section 2 and even taking into account highly non-canonical inverse systems,
direct/inverse systems represent a very restricted phenomenon only attested in a minority of
languages. Givón (1994) has proposed to extend this term to a broader set of phenomena.
Givón (1994, 9) defines the inverse as a type of de-transitive voice where ‘the patient is
more topical than the agent but the agent retains considerable topicality’, unlike the passive
where the agent is extremely non-topical (demoted or suppressed). He strictly distinguishes
between pragmatic inverse (the optional use of the inverse to indicate that the agent is more
topical than the patient) and semantic inverse (the obligatory use of the inverse conditioned
by the relative ranking of the agent and patient on a particular referential hierarchy).
Under this view, inverse constructions are not limited to hierarchical systems, but the concept can be extended to describe topic fronting, a phenomenon he designates as word-order
inverse. This extension of the concept of direct/inverse, while it has some merits in terms
of cross-linguistic comparison, may run into difficulties when analyzing the syntax of languages with a near-canonical direct/inverse system, as those languages also have topic
fronting and right extraposition, and topic fronting of the patient is not always correlated
with the use of the inverse.
Regardless of terminological preferences, Givonian text counts (in particular of referential
distance and topic persistence, Givón 1994, 10) have provided an easily reproductible and insightful basis on which to build a cross-linguistic study of direct/inverse systems. It has been
successfully applied to various languages with direct/inverse, such as Kutenai (Dryer 1994)
and Japhug Rgyalrong (Jacques 2010).
© 2014 The Authors
Language and Linguistics Compass © 2014 John Wiley & Sons Ltd
Language and Linguistics Compass (2014): 1–17, 10.1111/lnc3.12079
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
© 2014 The Authors
Language and Linguistics Compass © 2014 John Wiley & Sons Ltd
Table 8. Direct/inverse and other typological features.
Family
Zbu
Cree
Sahaptin
Navajo
Dargwa
Movima
Mapudungun
Head- Discontinuous
marking constituents
x
x
x
x
x
Nonlocal
x
x
x
Mixed Local
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
Case
Noun/verb
Ergative Strong
Obviative Strong
Tripartite Strong
Strong
Ergative Strong
Omnipredicative
Strong
Incorporation
Polypersonal marking
on the verb
x
x
x
x
x
x
x
x
Direct/Inverse Systems
11
Language and Linguistics Compass (2014): 1–17, 10.1111/lnc3.12079
Rgyalrongic (ST)
Algonquian (Algic)
Sahaptian
Athabaskan
NakhDaghestanian
Isolate
Isolate
Language
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
12 Guillaume Jacques and Anton Antonov
5. Diachronic Perspective
Recent work such as Cristofaro (2013) has called into question the functional validity of
referential hierarchies as an explanatory tool to account for the existence of direct/inverse
systems. Instead, it suggests that the structure of direct/inverse systems might be better
explained by taking into account the diachronic origin of these markers.
In this section, we provide a panchronic overview of direct and inverse markers: their
origins from and their development into other markers. Still, this type of research is
fraught with difficulties for two main reasons. First, most languages with direct/inverse
marking (such as Movima, Kutenai, Mapudungun) are isolates and thus cannot be easily
studied in a diachronic perspective. Second, since there are no languages with a nearcanonical inverse with a written tradition predating the seventeenth century (when
Wampanoag was first put to writing by John Eliot), the data presented here are reconstructions obtained by using the comparative method and may be subject to revision in
the future.
5.1. ORIGINS OF INVERSE MARKING
The best attested origin of inverse markers is from an earlier cislocative. Two additional
sources (an earlier passive or a third person marker) have been identified, but examples are
fewer and more difficult to interpret.
5.1.1. Cislocative
The term ‘cislocative’ is used to refer to markers expressing a motion towards the speaker,
both directional (‘verb hither’) and associated motion (‘come to verb’) ones.10
The inverse marker in Nez Percé (Sahaptian) used in local scenarios with second person
acting on first person has grammaticalized from an earlier cislocative marker, reconstructed
for Proto-Sahaptian as *-im (Rude 1997, 122).
(16) héexn-e
see-PST
I saw you.
(17) héxn-ím-e
see-CIS-PST
You saw me.
Interestingly, this marker has also grammaticalized into the ergative case suffix which
appears on the non-SAP agent in mixed scenarios in cases where one would expect inverse
marking on the verb (Rude 1997, 121-2).
3S/A→SAP-see-PST man-ERG
The man saw me/you. (NEZ PERCÉ)
This grammaticalization path is also attested in Sino-Tibetan. DeLancey (2013, 13-14)
reports that in several Kuki-Chin languages, a verb meaning ‘to come’ (reconstructed for
Proto-Kuki-Chin as *huŋ by VanBik 2009, 191) has grammaticalized to a varying degree
in different languages into a cislocative ‘hither’ prefix on motion verbs. Interestingly, in some
Northern and Central languages, it has developed the additional function of optionally
marking some transitive or ditransitive configurations involving a 1st person patient as in
Bawm (example 19) or 1st or 2nd person patient as in Sizang (example 20).
© 2014 The Authors
Language and Linguistics Compass © 2014 John Wiley & Sons Ltd
Language and Linguistics Compass (2014): 1–17, 10.1111/lnc3.12079
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
Direct/Inverse Systems
13
1SG father ERG cup CIS 1SG give
My father gives me the cup. (Reichle 1981, 147-149, cited in DeLancey 2013, 14)
you 1SG-CIS eat will FIN
I will eat you. (Stern 1984, 48, cited in DeLancey 2013, 13)
Even if we are not dealing with a fully grammaticalized direct/inverse system, the process
towards its shaping is well under way.
Similar developments which have not (as yet) reached the same target are also attested in
the unrelated languages Shasta (Shastan, extinct); the Northern Iroquoian languages Cayuga,
Tuscarora, and Mohawk (Mithun 1996); and also in Modern Japanese (Shibatani 2003, 2006;
Koga and Ohori 2008) and Kabardian (Kumakhov and Vamling 2006, 68ff). In all of these
languages, a cislocative has evolved, or is currently evolving, into a first person patient or a
portmanteau 2→1 marker.
As for the grammaticalization path involved, it need not have been the same in all of the
languages concerned. Based on data from Japanese (Japonic) and Circassian (NW Caucasian),
the following tentative grammaticalization pathway could be proposed:
(21)
COME
→
CISLOCATIVE associatedmotion
directional
→ (‘sudden action’)? → INVERSE
Indeed, apart from their incipient use as a marker of grammatical relations in mixed scenarios,
cislocative markers in these two languages have independently developed a secondary meaning
of ‘sudden action’, a semantic extension not attested in the other languages of the sample.
Therefore, the pathway proposed in (21) may not be applicable to other languages where the
inverse originated from an earlier cislocative marker, and so further research is needed.
5.1.2. Passive
A second grammaticalization pathway from passive to inverse has been proposed by several
authors, the most prominent of whom is Givón (1994).
This pathway could perhaps be attested in the case of Algonquian. At present, there is no
consensus on the origin of the inverse marker (used in the independent order) which has been
reconstructed for Proto-Algonquian as *-ekw (Bloomfield 1946, 98; Goddard 1979, 89).
Goddard (1974) pointed out a possible Wiyot cognate, but the data are too fragmentary to
draw any firm conclusions. Yurok, the only other relative of Algonquian, has a passive suffix
-oy (Robins 1958, 56-8), which presents affinities with inverse marking: it is required in
some 3→SAP forms and blocked in SAP→3. However, this passive suffix is not relatable
to Algonquian *-ekw by known phonetic laws (A. Garrett, p.c.).
Still, McLean (2001) has proposed that the Algonquian inverse suffix *-ekw originated
from a passive marker whose function was similar to the suffix -oy in Yurok. There is some
evidence for this in modern languages such as Cree and Ojibwe which still have a passive
construction distinct from the inverse one but sharing with it the *-ekw element. In Ojibwe
(Valentine 2001, 689), for instance, the passive SAP form is marked by -igoo, a compound
suffix which could be tentatively analyzed as resulting from the fusion of the inverse -igw
with the intransitive animate verb derivational suffix -aa.
5.1.3. Third Person
In still other languages, we observe a formal similarity between some inverse or direct markers
and various types of third person markers including agent, patient, or possessive affixes.
© 2014 The Authors
Language and Linguistics Compass © 2014 John Wiley & Sons Ltd
Language and Linguistics Compass (2014): 1–17, 10.1111/lnc3.12079
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
14 Guillaume Jacques and Anton Antonov
The clearest case is the inverse prefix found in Sino-Tibetan (Rgyalrongic and Kiranti)
languages. As first noticed by DeLancey (1981b), the Situ Rgyalrong third person possessive
prefix wə- is formally identical to the inverse marker. This is also true of other Rgyalrong
languages and of some Kiranti languages that have an inverse marker, such as Bantawa
(see Table 9, data from Gong and Doornenbal 2009).
T9
The similarity between the two sets of prefixes is striking and suggests a grammaticalization
from a third person marker into an inverse marker. While the exact pathway remains unclear
and thus requires further investigation, it is possible that non-finite verb forms carrying a third
person possessive prefix were reanalyzed as finite ones. In the case of Sino-Tibetan languages,
this scenario probably occurred in the ancestor of both Rgyalrongic and Kiranti languages
(see Jacques 2012a and DeLancey 2011).
5.2. STABILITY OF DIRECT/INVERSE SYSTEMS
Direct/inverse marking appears to be quite stable in at least one family: Algonquian. The
system can be reconstructed back to Proto-Algonquian (Goddard 1979, 89) and has been
preserved with little modification in all daughter languages.
However, in Sino-Tibetan, the only other large family where direct/inverse systems are
attested, near-canonical direct/inverse systems are only present in the Rgyalrong languages
described in Section 2.1. Other Sino-Tibetan languages show highly non-canonical or even
completely opaque inverse systems (cf. Kiranti languages described in Section 3.2).
In this connection, Lavrung (Huang 2007, 69-70; Lai 2013) shows an interesting evolution
which might help us understand how such opaque systems develop. A comparison of the data
in Table 10 with the Zbu Rgyalrong paradigm in 3 shows that the ə- prefix neatly corre- T10
sponds to the inverse wə- in Zbu and indeed the system we have in Lavrung, though slightly
less canonical than that of Zbu, is still what we would call a near-canonical inverse system.
Table 9. Comparison of inverse and third person prefixes in Rgyalrong and Kiranti.
Rgyalrong
Kiranti
Language
Inverse marker
Third person possessive
Situ
Tshobdun
Zbu
Bantawa
wəowə-i - < *u-
wəowə-i - < *u-
Table 10. Lavrung verbal paradigm (singular and plural forms).
© 2014 The Authors
Language and Linguistics Compass © 2014 John Wiley & Sons Ltd
Language and Linguistics Compass (2014): 1–17, 10.1111/lnc3.12079
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
Direct/Inverse Systems
15
Lavrung appears to differ from Zbu mainly in having lost direct/inverse contrast in nonlocal forms (3→3′ and 3′→3) to the benefit of the inverse form. This evolution appears
somewhat contrary to a priori expectation, given that inverse could be construed as being
the marked form and the direct the unmarked form. Lavrung thus represents a first step in
the direction of an opaque (direct/inverse) system. The prefix ə- cannot be analyzed synchronically as a third person agent marker, since it occurs in 2→1 forms, but it is not a canonical inverse marker anymore either, since it also occurs in all 3→3 forms.
There is no consensus as to how exactly the opaque systems observed in Kiranti came into
existence, but a possibility is that they originated from a direct/inverse system similar to that
observed in Rgyalrong languages, and then underwent a series of changes which thoroughly
reshaped the system starting with a generalization of the inverse marker in non-local scenarios as in the case of Lavrung. This in turn increased morphological opacity up to a point
where many individual morphemes (like Khaling i-) ceased to have a synchronically
straightforward morphosyntactic function.
6. Conclusion
Direct/inverse systems in the narrow sense are a rare type of agreement system, observed mainly
in languages of the Americas and the greater Himalayan region. Nevertheless, their theoretical
importance cannot be overestimated. Indeed, they have served as one of the models for the
development of the concept of referential hierarchies and the idea of setting hierarchical alignment
as a category distinct from accusative, ergative, active, or tripartite alignment.
Direct/inverse systems are attested in extremely endangered languages, and much work is
still needed for a detailed and meticulous description of such systems. In particular, the precise conditions for the use of direct and inverse forms in the non-local domain is a topic of
research, which cannot be based on elicitation, but requires extensive work on authentic oral
corpora. A finer-grained approach to morphosyntactic diversity is necessary, and recent research on Rgyalrongic languages for instance has revealed subtle but important differences
in the use of the inverse across closely related language varieties.
Acknowledgement
We would like to thank Sonia Cristofaro, Lynn Drapeau, Ives Goddard, Katharina Haude,
Enrique Palancar, Géraldine Walther, and two anonymous reviewers for useful comments
on earlier versions of this article. We are responsible for any remaining errors. This research
was funded by the HimalCo project (ANR-12-CORP-0006) and is related to the research
strand LR-4.11 ‘Automatic paradigm generation and language description’ of the Labex EFL
(funded by the ANR/CGI).
Short Biographies
Guillaume Jacques is a researcher at the CNRS (CRLAO) and specializes in language documentation, historical linguistics, and linguistic typology. He works mainly on Rgyalrongic
(Japhug Rgyalrong, Rtau) and Kiranti (Khaling) languages but has also done fieldwork on
Tibetan languages (Cone, Kami), Pumi, and various other Rgyalrongic languages. He has
published in journals such as Lingua, Linguistic Typology, and Diachronica.
Anton Antonov is an assistant professor of Japanese language and linguistics at INALCO (Paris).
He specializes in linguistic typology and historical linguistics, focusing on ‘Altaic’ languages as well as
Basque, Romance, and North-East Caucasian languages. He has published in Studies in Language
Companion Series. He is also currently involved in the documentation of the Rtau language.
© 2014 The Authors
Language and Linguistics Compass © 2014 John Wiley & Sons Ltd
Language and Linguistics Compass (2014): 1–17, 10.1111/lnc3.12079
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
16 Guillaume Jacques and Anton Antonov
Notes
* Correspondence address: Guillaume Jacques, CNRS-INALCO-EHESS, CRLAO, 2 Rue de Lille, 75007 Paris,
Q2
France. E-mail: rgyalrongskad@gmail.com
1
Indexation systems involving three arguments exist but are rarer and will not be considered in this section.
The same applies in languages with clusivity (a distinction between first person inclusive vs exclusive) to the combination of first inclusive and second person.
3
See Haspelmath (2011) for a discussion of the terms S (only argument of an intransitive verb), A (more agent-like
argument of a transitive verb), and O (more patient-like argument of a transitive verb). In terms of Bickel et al. (, 9)’s
approach, all persons trigger agreement in all three syntactic roles.
4
Direct/inverse systems are often illustrated with data from Algonquian languages, but they generally lack direct/
inverse marking in the local domain and are thus farther removed from the canon than Rgyalrong languages.
5
For an overview of referential or prominence hierarchies, see Lockwood and Macauley (2012). These hierarchies,
initially proposed in order to account for particular splits in case patterns (see Silverstein 1976), have also been used to
explain a variety of other phenomena, including the distribution of direct and inverse markers and slot accessibility.
However, it is not always possible to account for all phenomena within a given language by positing a single hierarchy,
as Zúñiga 2006 has shown. In Cree for instance, no less than four distinct hierarchies are needed to explain direct/inverse
marking and the allocation of person markers to the correct slot.
6
Hierarchies involving 3 > SAP may govern slot accessibility, as in the case of slots 5-7 in Cree which follow the
hierarchy 1PL > 1INCL/2PL > 3 animate > SAP.SG > 3 inanimate, Zúñiga 2006, 86. This type of hierarchy has however
never been described for direct/inverse morphemes.
7
It is generally considered that the second person outranks the first person (2 > 1) in Algonquian languages, but this
refers to a distinct hierarchy related to the slot accessibility of person prefixes, not the distribution of direct and inverse
forms. Concerning obviative inanimates, see a recent study by Muehlbauer (2012).
8
The first case is represented by languages with proximate/obviative marking only in the non-local domain (Kutenai,
Navajo) and languages with inverse only in local forms such as Nez Percé.
9
The Tupi-Guarani family provides many examples of such systems (see for instance Rose 2009).
10
On the grammaticalization of associated motion markers from motion verbs, see Jacques (2013).
2
Works Cited
Aissen, Judith. 1997. On the syntax of obviation. Language 73.4. 705–50.
Bickel, Balthasar, Giorgio Iemmolo, Taras Zakharko, and Alena Witzlack-Makarevich. in press. Patterns of alignment in
verb agreement. Languages across boundaries: studies in memory of Anna Siewierska, ed. by D. Bakker and M.
Haspelmath. Berlin: de Gruyter Mouton.
Bloomfield, Leonard. 1946. Algonquian. Linguistic structures of native America, ed. by H. Hoijer and C. Osgood, volume 6 of Viking fund publications in anthropology, 85–129. New York: Wenner-Gren Foundation.
Corbett, Greville G. 2007. Canonical typology, suppletion and possible words. Language 83.1. 8–42.
Cristofaro, Sonia. 2013. The referential hierarchy: reviewing the evidence in diachronic perspective. Languages across
boundaries, ed. by D. Bakker and M. Haspelmath, 69–94. Berlin, New York: Mouton de Gruyter.
DeLancey, Scott. 1981a. An interpretation of split ergativity. Language 57.3. 626–57.
——. 1981b. The category of direction in Tibeto-Burman. Linguistics of the Tibeto-Burman Area 6.1. 83–101.
——. 2011. Notes on verb agreement prefixes in Tibeto-Burman. Himalayan Linguistics Journal 10.1. 1–29.
——. 2013. Argument indexation (verb agreement) in Kuki-Chin. Retrieved on 7 August 2013 from: http://www.
academia.edu/4155394/Argument_Indexation_Verb_Agreement_in_Kuki-Chin, paper presented at the Workshop
on Issues in Kuki-Chin Linguistics at the 46th International Conference on Sino-Tibetan Languages and Linguistics,
Dartmouth College.
Doornenbal, Marius. 2009. A grammar of Bantawa: grammar, paradigm tables, glossary and texts of a Rai language of
Eastern Nepal. Doctoral Dissertation, Leiden University.
van Driem, George. 1993. The Proto-Tibeto-Burman verbal agreement system. Bulletin of the School of Oriental and
African Studies 61.2. 292–334.
Dryer, Matthew. 1994. The discourse function of the Kutenai inverse. Voice and inversion, ed. by T. Givon, 65–99.
Amsterdam: Benjamins.
Givón, Talmy. 1994. The pragmatics of de-transitive voice: functional and typological aspects of inversion. Voice and
Q4
inversion. Amsterdam: Benjamins.
Goddard, Ives. 1974. Remarks on the Algonquian independent indicative. International Journal of American Linguistics
40.4. 317–27.
© 2014 The Authors
Language and Linguistics Compass © 2014 John Wiley & Sons Ltd
Language and Linguistics Compass (2014): 1–17, 10.1111/lnc3.12079
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
Direct/Inverse Systems
17
——. 1979. Comparative Algonquian. The languages of native America, ed. by L. Campbell and M. Mithun, 70–132.
Austin: University of Texas Press.
Gong, Xun. forthcoming. Personal agreement system of Zbu rGyalrong (Ngyaltsu variety). Transactions of the Philological Society.
Haspelmath, Martin. 2011. On S, A, P, T, and R as comparative concepts for alignment typology. Linguistic Typology
15.3. 535–67.
Haude, Katharina. 2009. Hierarchical alignment in Movima. International Journal of American Linguistics 75.4. 513–32.
Huang, Bufan. 2007. Lawurongyu yanjiu
(study on the Lavrung language). Beijing: Minzu
chubanshe.
Jacques, Guillaume. 2009. The origin of vowel alternations in the Tangut verb. Language and Linguistics 10.1. 17–28.
——. 2010. The inverse in Japhug Rgyalrong. Language and Linguistics 11.1. 127–57.
——. 2012a. Agreement morphology: the case of Rgyalrongic and Kiranti. Language and Linguistics 13.1. 83–116.
——. 2012b. Argument demotion in Japhug Rgyalrong. Ergativity, valency and voice, ed. by K. Haude and G. Authier,
199–226. Berlin: Mouton De Gruyter.
——. 2013. Harmonization and disharmonization of affix ordering and basic word order. Linguistic Typology 17.2.
187–217.
Jacques, Guillaume, Aimée Lahaussois, Boyd Michailovsky, and Dhan Bahadur Rai. 2012. An overview of Khaling
verbal morphology. Language and Linguistics 13.6. 1095–170.
Kepping, Ksenija Borisovna. 1981. Agreement in the Tangut verb. Linguistics of the Tibeto-Burman Area 6.1. 39–48.
Klaiman, M. H. 1991. Grammatical voice. Cambridge: Cambridge University Press.
——. 1992. Inverse languages. Lingua 88.3/4. 227–61.
Koga, Hiroaki, and Toshio Ohori. 2008. Reintroducing inverse constructions in Japanese. Investigations of the syntaxsemantics-pragmatics interface, ed. by R. D. Van Valin Jr. Amsterdam: John Benjamins.
Kumakhov, Mukhadin, and Karina Vamling. 2006. Èrgativnost’ v čerkesskix jazykax. Malmo University.
Lai, Yunfan. 2013. La morphologie affixale du lavrung wobzi. Master’s thesis, Université Paris III.
Lockwood, Hunter T., and Monica Macauley. 2012. Prominence hierarchies. Language and Linguistics Compass 6.7.
431–46.
McLean, Lisa M. 2001. A passive to inverse reanalysis in Cree. Master’s thesis, University of Manitoba at Winipeg.
Mithun, Marianne. 1996. New directions in referentiality. Studies in anaphora, ed. by B. Fox, volume 33 of Typological
studies in language, 413–35. Amsterdam: John Benjamins.
Muehlbauer, Jeffrey. 2012. the relation of switch-reference, animacy, and obviation in Plains Cree. International Journal
of American Linguistics 78.2. 203–38.
Nichols, Johanna. 1992. Linguistic diversity in space and time. Chicago: University of Chicago Press.
Reichle, Verena. 1981. Bawm language and lore. Peter Lang.
Robins, R. H. 1958. The Yurok language. Berkeley: University of California Press.
Rose, Françoise. 2009. A hierarchical indexation system: the example of Emerillon (Teko). New challenges in typology,
ed. by P. Epps and A. Arkhipov, 63–83. Berlin: Mouton de Gruyter.
Rude, Noel. 1997. On the history of nominal case in Sahaptian. International Journal of American Linguistics 63. 113–43.
Shibatani, Masayoshi. 2003. Directional verbs in Japanese. Motion, direction and location in languages: in honor of
Zygmunt Frajzyngier, ed. by E. Shay and U. Seibert, 259–86. John Benjamins.
——. 2006. On the conceptual framework for voice phenomena. Linguistics 44. 217–69.
Siewierska, Anna. 1998. On nominal and verbal person marking. Linguistic Typology 2. 1–56.
Silverstein, Michael. 1976. Hierarchy of features and ergativity. Grammatical categories in Australian languages, ed. by
R. M. W. Dixon. Canberra: Australian Institute of Aboriginal Studies.
Stern, Theodore. 1984. Sizang (Siyin) Chin texts. Linguistics of the Tibeto-Burman Area 8. 43–58.
Sumbatova, Nina R., and Rasul O. Mutalov. 2003. A grammar of Icari Dargwa . München: Lincom Europa.
Sun, Jackson T.-S., and Shidanluo. 2002. Caodeng Jiarongyu yu rentong dengdi xiangguan de yufa xianxiang
(empathy hierarchy in Caodeng rGyalrong grammar).
Language and Linguistics 3.1. 79–99.
Valentine, J. Randolph. 2001. Nishnaabemwin reference grammar. Toronto: University of Toronto Press.
VanBik, Kenneth. 2009. Proto-Kuki-Chin, volume 8 of STEDT monograph series. University of California.
Willie, Mary-Ann. 2000. The inverse voice and possessive yi/bi. International Journal of American Linguistics 66.3.
360–82.
Wolfart, H. Christoph. 1973. Plains Cree: a grammatical study. Transactions of the American Philosophical Society
63.5. 1–90.
Zúñiga, Fernando. 2006. Deixis and alignment – inverse systems in indigenous languages of the Americas. Amsterdam:
Benjamins.
© 2014 The Authors
Language and Linguistics Compass © 2014 John Wiley & Sons Ltd
Language and Linguistics Compass (2014): 1–17, 10.1111/lnc3.12079
Q5
Q6
Q7
Q8
Q9
Q10
Author Query Form
Journal: Language and Linguistics Compass
Article: lnc_312079
Dear Author,
During the copyediting of your paper, the following queries arose. Please respond to these by annotating your proofs with the necessary changes/additions.
• If you intend to annotate your proof electronically, please refer to the E-annotation guidelines.
• If you intend to annotate your proof by means of hard-copy mark-up, please refer to the proof
mark-up symbols guidelines. If manually writing corrections on your proof and returning it by
fax, do not write too close to the edge of the paper. Please remember that illegible mark-ups
may delay publication.
Whether you opt for hard-copy or electronic annotation of your proofs, we recommend that you provide additional clarification of answers to queries by entering your answers on the query sheet, in
addition to the text mark-up.
Query No.
Query
Q1
AUTHOR: Please provide an abstract.
Q2
AUTHOR: Please check address provided for the affiliation
in the correspondence field.
Q3
AUTHOR: Tables have been renumbered to obtain a
sequential order.Tables 3 has been changed to Table 4 and
vice versa. Please check if correct.
Q4
AUTHOR: Please provide the Editor(s) and page number for
Givón, 1994.
Q5
AUTHOR: If Gong, to appear has now been published in
print, please add relevant volume/issue/page/year information.
Q6
AUTHOR: Please provide the city location of publisher for
Kumakhov and Vamling, 2006.
Q7
AUTHOR: Please provide the city location of publisher for
Reichle, 1981.
Q8
AUTHOR: Please provide the city location of publisher for
Shibatani, 2003.
Q9
AUTHOR: Please provide the initial(s) of Shidanluo.
Q10
AUTHOR: Please provide the city location of publisher for
VanBik, 2009.
Remark
USING e-ANNOTATION TOOLS FOR ELECTRONIC PROOF CORRECTION
Required software to e-Annotate PDFs: Adobe Acrobat Professional or Adobe Reader (version 7.0 or
above). (Note that this document uses screenshots from Adobe Reader X)
The latest version of Acrobat Reader can be downloaded for free at: http://get.adobe.com/uk/reader/
Once you have Acrobat Reader open on your computer, click on the Comment tab at the right of the toolbar:
This will open up a panel down the right side of the document. The majority of
tools you will use for annotating your proof will be in the Annotations section,
pictured opposite. We’ve picked out some of these tools below:
1. Replace (Ins) Tool – for replacing text.
2. Strikethrough (Del) Tool – for deleting text.
Strikes a line through text and opens up a text
box where replacement text can be entered.
How to use it
Highlight a word or sentence.
Click on the Replace (Ins) icon in the Annotations
section.
Strikes a red line through text that is to be
deleted.
How to use it
4. Add sticky note Tool – for making notes at
specific points in the text.
Highlights text in yellow and opens up a text
box where comments can be entered.
How to use it
Click on the Strikethrough (Del) icon in the
Annotations section.
Type the replacement text into the blue box that
appears.
3. Add note to text Tool – for highlighting a section
to be changed to bold or italic.
Highlight a word or sentence.
Marks a point in the proof where a comment
needs to be highlighted.
How to use it
Highlight the relevant section of text.
Click on the Add note to text icon in the
Annotations section.
Type instruction on what should be changed
regarding the text into the yellow box that
appears.
Click on the Add sticky note icon in the
Annotations section.
Click at the point in the proof where the comment
should be inserted.
Type the comment into the yellow box that
appears.
USING e-ANNOTATION TOOLS FOR ELECTRONIC PROOF CORRECTION
5. Attach File Tool – for inserting large amounts of
text or replacement figures.
6. Add stamp Tool – for approving a proof if no
corrections are required.
Inserts an icon linking to the attached file in the
appropriate pace in the text.
How to use it
Inserts a selected stamp onto an appropriate
place in the proof.
How to use it
Click on the Attach File icon in the Annotations
section.
Click on the proof to where you’d like the attached
file to be linked.
Select the file to be attached from your computer
or network.
Select the colour and type of icon that will appear
in the proof. Click OK.
Click on the Add stamp icon in the Annotations
section.
Select the stamp you want to use. (The Approved
stamp is usually available directly in the menu that
appears).
Click on the proof where you’d like the stamp to
appear. (Where a proof is to be approved as it is,
this would normally be on the first page).
7. Drawing Markups Tools – for drawing shapes, lines and freeform
annotations on proofs and commenting on these marks.
Allows shapes, lines and freeform annotations to be drawn on proofs and for
comment to be made on these marks..
How to use it
Click on one of the shapes in the Drawing
Markups section.
Click on the proof at the relevant point and
draw the selected shape with the cursor.
To add a comment to the drawn shape,
move the cursor over the shape until an
arrowhead appears.
Double click on the shape and type any
text in the red box that appears.
For further information on how to annotate proofs, click on the Help menu to reveal a list of further options: