05.04.2013 Views

1 Low applicatives and the mapping hypothesis in Sumerian J. Cale ...

1 Low applicatives and the mapping hypothesis in Sumerian J. Cale ...

1 Low applicatives and the mapping hypothesis in Sumerian J. Cale ...

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

<strong>Low</strong> <strong>applicatives</strong> <strong>and</strong> <strong>the</strong> <strong>mapp<strong>in</strong>g</strong> hypo<strong>the</strong>sis <strong>in</strong> <strong>Sumerian</strong><br />

J. <strong>Cale</strong> Johnson<br />

Annual Meet<strong>in</strong>g of <strong>the</strong> L<strong>in</strong>guistic Society of America<br />

(January 7, 2005, Syntax <strong>and</strong> Morphology panel, 2pm)<br />

[[Introduction]]<br />

<strong>Sumerian</strong>, <strong>the</strong> first written language, which is preserved on clay tablets dat<strong>in</strong>g to four to five<br />

thous<strong>and</strong> years ago, is a radically SOV language that seems to exhibit almost no movement <strong>in</strong> its<br />

syntax. What it apparently lacks <strong>in</strong> movement is more than made up for through a vast <strong>and</strong><br />

complex morphology consist<strong>in</strong>g of a full set of case-mark<strong>in</strong>g postpositions as well as an<br />

<strong>in</strong>ord<strong>in</strong>ately complex verb with from twelve to fourteen paradigmatic slots, all but two of which<br />

are prefixes.<br />

What I am here today to report on is <strong>the</strong> role of possessor rais<strong>in</strong>g <strong>and</strong> <strong>the</strong> non-specificity of <strong>the</strong><br />

possessed entity <strong>in</strong> possessor-rais<strong>in</strong>g constructions <strong>in</strong> <strong>the</strong> formation of a small set of noun-verb<br />

complex predicate constructions <strong>in</strong> <strong>Sumerian</strong>. One type of such noun-verb complex predicate fits<br />

very nicely <strong>in</strong>to Pylkkänen’s description of a low source applicative, <strong>in</strong> which a <strong>the</strong>me is<br />

transferred from one possessor to ano<strong>the</strong>r. The crucial po<strong>in</strong>t, however, is that <strong>the</strong>re is ano<strong>the</strong>r<br />

type of noun-verb complex that is morphologically identical with <strong>the</strong> <strong>Sumerian</strong> low applicative<br />

but that is causative <strong>in</strong> mean<strong>in</strong>g. Thus I argue <strong>in</strong> <strong>the</strong> follow<strong>in</strong>g that <strong>the</strong>se two constructions are <strong>in</strong><br />

fact morphologically identical but differ <strong>in</strong> <strong>the</strong> syntactic position of <strong>the</strong> nom<strong>in</strong>al component of<br />

<strong>the</strong> complex predicate: where this nom<strong>in</strong>al component is non-specific <strong>and</strong> rema<strong>in</strong>s with<strong>in</strong> <strong>the</strong> VP,<br />

it undergoes existential closure <strong>and</strong> restricts <strong>the</strong> predicate along <strong>the</strong> l<strong>in</strong>es sketched by ei<strong>the</strong>r<br />

1


Chung <strong>and</strong> McCloskey (2004) or Van Geenhoven <strong>and</strong> McNally (2005). When, however, <strong>the</strong><br />

nom<strong>in</strong>al component is specific <strong>and</strong> moves out of <strong>the</strong> VP, it forms a k<strong>in</strong>d of <strong>in</strong>ternally headed<br />

relative that is causative <strong>in</strong> mean<strong>in</strong>g. In <strong>the</strong> end, I use a negative contrastive focus construction <strong>in</strong><br />

comb<strong>in</strong>ation with Dies<strong>in</strong>g’s <strong>mapp<strong>in</strong>g</strong> hypo<strong>the</strong>sis to differentiate <strong>the</strong> applicative <strong>and</strong> <strong>the</strong> causative<br />

<strong>in</strong> <strong>Sumerian</strong>.<br />

[[Possessor rais<strong>in</strong>g]]<br />

In a language like English, possessor rais<strong>in</strong>g is a relatively limited phenomenon found <strong>in</strong><br />

constructions <strong>in</strong> which <strong>the</strong> possessed entity—ra<strong>the</strong>r than <strong>the</strong> possessor—has raised out of <strong>the</strong><br />

ord<strong>in</strong>ary, prenom<strong>in</strong>al possessive DP to form an encompass<strong>in</strong>g NP as <strong>in</strong> Kayne’s famous example<br />

<strong>in</strong> (1): “two pictures of John’s.” Kayne is primarily <strong>in</strong>terested <strong>in</strong> what “of” is do<strong>in</strong>g <strong>in</strong> this<br />

construction, <strong>the</strong> parallels with datives of possession <strong>in</strong> o<strong>the</strong>r languages <strong>and</strong> how this feeds <strong>in</strong>to<br />

his <strong>the</strong>ory of relative clause formation, but <strong>the</strong> reason why this example is important to me is that<br />

it offers a mechanism for attach<strong>in</strong>g a non-specific NP like “two pictures” to a highly specific DP<br />

like <strong>the</strong> possessor <strong>in</strong> (1). The non-specificity <strong>and</strong> consequent absence of a DP superstructure is<br />

particularly clear <strong>in</strong> an existential sentence such as <strong>in</strong> (2). Note that if <strong>the</strong> unraised “John’s two<br />

pictures” were to appear <strong>in</strong> (2), it would yield a pair-list read<strong>in</strong>g, <strong>the</strong>reby demonstrat<strong>in</strong>g <strong>the</strong><br />

specificity contrast.<br />

Now, <strong>in</strong> <strong>the</strong> k<strong>in</strong>d of language where possessor rais<strong>in</strong>g is typically described—<strong>and</strong> here I’m<br />

th<strong>in</strong>k<strong>in</strong>g of primarily of Biblical Hebrew ra<strong>the</strong>r than <strong>the</strong> Hungarian that has figured so centrally<br />

<strong>in</strong> syntactic research—it is actually <strong>the</strong> possessor that raises ra<strong>the</strong>r than <strong>the</strong> possessed entity: <strong>in</strong><br />

fact <strong>the</strong> possessor raises all <strong>the</strong> way out of <strong>the</strong> DP that it specifies <strong>and</strong> forms a separate phrase <strong>in</strong><br />

2


<strong>the</strong> dative case as <strong>in</strong> <strong>the</strong> example <strong>in</strong> (3) from Biblical Hebrew: rΩ’îtî ben le-yiåΩ y . Hebrew splits<br />

<strong>the</strong> possessor off <strong>in</strong>to a separate phrase <strong>in</strong> <strong>the</strong> dative case, while English buries <strong>the</strong> possessor<br />

deep with<strong>in</strong> whatever k<strong>in</strong>d of phrase is headed by “of” <strong>in</strong> “two pictures of John’s.” Both<br />

mechanisms manage to separate a non-specific possessed noun from a specific or even def<strong>in</strong>ite<br />

possessor <strong>and</strong> prevent <strong>the</strong> transmission of specificity or def<strong>in</strong>iteness from possessor to possessed<br />

entity.<br />

In 1999, Gábor Zólyomi was <strong>the</strong> first person to identify possessor rais<strong>in</strong>g <strong>in</strong> <strong>Sumerian</strong> as <strong>in</strong> (4).<br />

In (4), <strong>the</strong> raised possessor is a deity named N<strong>in</strong>girsu—on <strong>the</strong> far left—while <strong>the</strong> possessed<br />

entity—åu mean<strong>in</strong>g ‘h<strong>and</strong>’—occurs immediately before <strong>the</strong> verb. S<strong>in</strong>ce <strong>the</strong> two components of<br />

<strong>the</strong> raised possessor construction are not adjacent, it isn’t immediately clear whe<strong>the</strong>r <strong>the</strong> phrase<br />

<strong>in</strong> <strong>the</strong> dative case marks <strong>the</strong> ord<strong>in</strong>ary dative of a recipient, or alternatively, <strong>the</strong> raised possessor<br />

of åu <strong>and</strong> only <strong>in</strong>directly a recipient. As already suggested by Zólyomi, <strong>the</strong> syntactic problem<br />

that possessor rais<strong>in</strong>g solves <strong>in</strong> this sentence is that <strong>the</strong>re are simply too many arguments: four <strong>in</strong><br />

total. The ergative agent, a k<strong>in</strong>g named Eannatum, is unexpressed; <strong>the</strong> absolutive patient, aåa<br />

ki’aÑani—mean<strong>in</strong>g ‘his beloved field’—is <strong>in</strong> <strong>the</strong> middle of <strong>the</strong> clause; åu-ni-a—literally ‘<strong>in</strong> his<br />

h<strong>and</strong>’—immediately before <strong>the</strong> verb, acts as <strong>the</strong> third <strong>and</strong> what would normally be <strong>the</strong> f<strong>in</strong>al goal<br />

argument. The problem is that <strong>the</strong> predicate <strong>in</strong> (4) is a complex predicate made up of <strong>the</strong> noun åu<br />

<strong>and</strong> <strong>the</strong> verbal root gi, mean<strong>in</strong>g as a whole ‘to return someth<strong>in</strong>g to someone’. S<strong>in</strong>ce åu-ni-a<br />

already occupies <strong>the</strong> goal argument, <strong>the</strong> recipient is coded as <strong>the</strong> possessor of åu. The ra<strong>the</strong>r<br />

“<strong>in</strong>volved” nature of <strong>the</strong> construction is somewhat clarified by comparison with an example of<br />

<strong>the</strong> same construction <strong>in</strong> which <strong>the</strong> recipient rema<strong>in</strong>s unexpressed. Here <strong>in</strong> (5), which I have<br />

modified slightly from <strong>the</strong> orig<strong>in</strong>al, <strong>the</strong> *bi- prefix is used <strong>in</strong> place of <strong>the</strong> *-ni- prefix seen <strong>in</strong> (4)<br />

3


<strong>and</strong>, more importantly, both <strong>the</strong> raised possessor <strong>and</strong> <strong>the</strong> resumptive pronoun are absent. The<br />

nom<strong>in</strong>al component of <strong>the</strong> verbal complex, åu, still occurs <strong>in</strong> <strong>the</strong> locative case as <strong>the</strong> idiom<br />

dem<strong>and</strong>s, but <strong>the</strong>re is no resumptive pronoun.<br />

[[The morphology of <strong>the</strong> verbal prefix]]<br />

The morphosyntactic test that differentiates a dative function<strong>in</strong>g as a raised possessor from <strong>the</strong><br />

dative associated with an ord<strong>in</strong>ary recipient is <strong>the</strong> form of <strong>the</strong> verbal prefix. If we compare <strong>the</strong><br />

verbal prefix <strong>in</strong> (4) to <strong>the</strong> prefix of an ord<strong>in</strong>ary ditransitive verb like <strong>the</strong> one <strong>in</strong> (6), it should be<br />

immediately apparent that <strong>the</strong> prefix <strong>in</strong> (6) is *-na-, <strong>the</strong> third person animate dative prefix.<br />

Unlike <strong>the</strong> *-ni- prefix, <strong>the</strong>re is widespread agreement that *-na- <strong>and</strong> <strong>the</strong> o<strong>the</strong>r members of its<br />

paradigm code <strong>the</strong> dative case, prototypically represent<strong>in</strong>g recipients, addressees <strong>and</strong><br />

benefactees.<br />

The *-ni- prefix, however, shows up <strong>in</strong> a variety of constructions, but crucially whenever<br />

possessor rais<strong>in</strong>g seems to be present, we also f<strong>in</strong>d *-ni- or one of <strong>the</strong> o<strong>the</strong>r members of its<br />

paradigm. This correlation is fur<strong>the</strong>r steng<strong>the</strong>ned by <strong>the</strong> morphological form of <strong>the</strong> *-ni-<br />

paradigm itself as <strong>in</strong> (7). One of <strong>the</strong> st<strong>and</strong>ard ways of mark<strong>in</strong>g a noun as def<strong>in</strong>ite or topical <strong>in</strong><br />

<strong>Sumerian</strong> is through <strong>the</strong> use of redundant possessive pronouns, so it makes a certa<strong>in</strong> amount of<br />

sense that non-specificity <strong>and</strong> <strong>the</strong> predicational quality often associated with non-specific<br />

nom<strong>in</strong>al phrases would be <strong>in</strong>dicated by form<strong>in</strong>g a raised possessor construction <strong>and</strong> at <strong>the</strong> same<br />

time—or perhaps as part of <strong>the</strong> same derivational or historical process—mov<strong>in</strong>g a possessive<br />

pronoun that agrees <strong>in</strong> person <strong>and</strong> gender with <strong>the</strong> possessor <strong>in</strong>to <strong>the</strong> verbal predicate itself <strong>in</strong> <strong>the</strong><br />

form of a prefix. The pronoun <strong>the</strong>n functions as a clause-level possessive predicate that takes a<br />

4


non-specific possessed entity immediately before <strong>the</strong> verb <strong>and</strong> codes <strong>the</strong> possessor <strong>in</strong> a separate<br />

postpositional phrase that tends to move to clause-<strong>in</strong>itial position where datives generally occur<br />

<strong>in</strong> <strong>Sumerian</strong>.<br />

[[Applicative/causative alternation]]<br />

Let me emphasize here that <strong>the</strong>re are many different k<strong>in</strong>ds of complex predicate <strong>in</strong> <strong>Sumerian</strong>, but<br />

<strong>the</strong> type that I have been work<strong>in</strong>g on recently is somewhat different from <strong>the</strong> rest <strong>in</strong> that <strong>the</strong><br />

nom<strong>in</strong>al component of <strong>the</strong> class of complex predicates that I would like to talk about never bears<br />

ei<strong>the</strong>r a resumptive pronoun or any case-mark<strong>in</strong>g postposition o<strong>the</strong>r than, perhaps, <strong>the</strong> zero-<br />

marked nom<strong>in</strong>ative-absolutive case. At <strong>the</strong> same time, certa<strong>in</strong> features of <strong>the</strong> nom<strong>in</strong>al component<br />

such as alienability or specificity seem to differentiate verbs of direct perception <strong>and</strong> deprivation<br />

from causatives.<br />

From a simplistic English po<strong>in</strong>t of view, <strong>the</strong> syntactic association between direct perception <strong>and</strong><br />

causation is fairly evident <strong>in</strong> that both constructions take small clause complements (“I saw<br />

[Mary leave]” vs. “I made [Mary leave]”). The same is true for several types of <strong>in</strong>f<strong>in</strong>itival<br />

complement <strong>in</strong> both Romance <strong>and</strong> Germanic. This was <strong>the</strong> approach <strong>in</strong> Kayne’s early work on<br />

French Syntax (Kayne 1975), nicely summarized <strong>in</strong> Guasti’s comparative study of Causative<br />

<strong>and</strong> Perception Verbs (Guasti 1993). Recent work by Pylkkänen (2000; 2002) <strong>and</strong> Harley<br />

(2002), build<strong>in</strong>g on Larson’s classic description of double object constructions (1988), has<br />

offered a slightly different elucidation of a set of similar constructions <strong>in</strong> terms of an<br />

applicative/causative alternation. The similarity between verbs of direct perception <strong>and</strong> causative<br />

verbs is particularly clear <strong>in</strong> <strong>Sumerian</strong>,where both a perceived entity <strong>and</strong> a causee appear <strong>in</strong> <strong>the</strong><br />

5


dative case as <strong>the</strong> raised possessor of an absolutive direct object that appears immediately before<br />

<strong>the</strong> verb <strong>and</strong> forms part of a verbal complex. Both constructions make use of <strong>the</strong> NI paradigm <strong>in</strong><br />

<strong>the</strong> verb ra<strong>the</strong>r than <strong>the</strong> ord<strong>in</strong>ary dative series.<br />

In (8), <strong>the</strong> verbal complex is made up of igi ‘eye’ <strong>and</strong> du—mean<strong>in</strong>g uncerta<strong>in</strong> but perhaps ‘to<br />

spread’ or ‘to hold’—while gig.e ‘wheat’ acts as <strong>the</strong> raised possessor of igi <strong>and</strong> codes <strong>the</strong><br />

perceived entity. The animacy, or better, <strong>the</strong> <strong>in</strong>animacy of <strong>the</strong> wheat is <strong>in</strong>dicated by <strong>the</strong> use of <strong>the</strong><br />

<strong>in</strong>animate dative postposition *-e <strong>and</strong> <strong>the</strong> *bi- prefix on <strong>the</strong> verb. In (9), we have <strong>the</strong> same<br />

construction except that <strong>the</strong> raised possessor—that is to say <strong>the</strong> perceived entity—is a deity <strong>and</strong><br />

hence animate. The possessor is coded <strong>the</strong>refore by <strong>the</strong> animate dative postposition *-ra, while<br />

<strong>the</strong> verbal prefix is *-ni-, <strong>the</strong> third person animate member of <strong>the</strong> NI paradigm. At least <strong>in</strong> terms<br />

of morphology <strong>and</strong> apparent surface distribution, <strong>the</strong> causative exemplars <strong>in</strong> (10) <strong>and</strong> (11) are<br />

<strong>the</strong> same as <strong>the</strong> applicative ones. In (10), a group of goats is <strong>the</strong> causee <strong>and</strong> is followed by <strong>the</strong><br />

<strong>in</strong>animate dative case, *-e, with <strong>the</strong> <strong>in</strong>animate *bi- prefix, while <strong>the</strong> animate causee <strong>in</strong> (11) takes<br />

<strong>the</strong> animate dative postposition, *-ra, <strong>and</strong> occurs with <strong>the</strong> *-ni- verbal prefix. In 2004, <strong>in</strong> my<br />

dissertation, I argued that <strong>the</strong> alienability of <strong>the</strong> nom<strong>in</strong>al component of <strong>the</strong> verbal complex<br />

determ<strong>in</strong>ed whe<strong>the</strong>r or not it was causative: <strong>in</strong>alienable nouns form verbs of perception, whereas<br />

alienable nouns form causatives. The problem with this view is that <strong>the</strong>re are simply too many<br />

exceptions, hence my appeal to specificity.<br />

[[<strong>Low</strong> source applicative <strong>in</strong> <strong>Sumerian</strong>]]<br />

The breakthrough, if you will . . . <strong>the</strong> big new idea that allows possessor rais<strong>in</strong>g to figure<br />

centrally <strong>in</strong> a typology of <strong>applicatives</strong> is Pylkkänen’s description of <strong>the</strong> low applicative (2002).<br />

6


Now, to be fair to Pylkkänen, she has argued quite extensively <strong>in</strong> opposition to, for example,<br />

Idan L<strong>and</strong>au’s suggestion that low applicative constructions are simply a form of possessor<br />

rais<strong>in</strong>g. In my view, however, <strong>the</strong> <strong>in</strong>teraction between <strong>applicatives</strong>, possessor rais<strong>in</strong>g <strong>and</strong><br />

specificity is still very much <strong>in</strong> flux. All that I would like to do today is throw ano<strong>the</strong>r batch of<br />

admittedly messy data <strong>in</strong>to <strong>the</strong> pot <strong>and</strong> leave it as a puzzle for <strong>the</strong> <strong>the</strong>oreticians. In simple terms,<br />

a low applicative is low because <strong>the</strong> po<strong>in</strong>t at which <strong>the</strong> applied argument—<strong>the</strong> argument that is<br />

be<strong>in</strong>g added <strong>in</strong>to <strong>the</strong> structure of <strong>the</strong> verb—attaches to <strong>the</strong> syntactic tree is below <strong>the</strong> level of <strong>the</strong><br />

VP. Thus <strong>the</strong>re is an <strong>in</strong>herent connection between Pylkkänen’s low applicative <strong>and</strong> Dies<strong>in</strong>g’s<br />

<strong>mapp<strong>in</strong>g</strong> hypo<strong>the</strong>sis, which also sees <strong>the</strong> VP as <strong>the</strong> po<strong>in</strong>t at which specific <strong>and</strong> non-specific<br />

nom<strong>in</strong>al phrases are differentiated. If you’ll compare <strong>the</strong> two Korean sentences <strong>in</strong> (12) <strong>and</strong> (13),<br />

note that <strong>the</strong> only difference between <strong>the</strong> two is that <strong>the</strong> possessor is coded by a genitive case <strong>in</strong><br />

(12), but by a raised possessor <strong>in</strong> (13), <strong>and</strong> it is <strong>the</strong> example <strong>in</strong>volv<strong>in</strong>g possessor rais<strong>in</strong>g that<br />

<strong>in</strong>cludes <strong>the</strong> notion that <strong>the</strong> possession of <strong>the</strong> r<strong>in</strong>g transfers from Mary to <strong>the</strong> thief. That is to<br />

say, Mary is <strong>the</strong> possessor of <strong>the</strong> r<strong>in</strong>g at its source, hence <strong>the</strong> term<strong>in</strong>ology of <strong>the</strong> example <strong>in</strong> (13)<br />

as a low source applicative: it is a low applicative because <strong>the</strong> applied argument, namely Mary,<br />

<strong>in</strong>itially attaches below <strong>the</strong> level of <strong>the</strong> VP <strong>and</strong> is <strong>the</strong>refore <strong>in</strong>volved <strong>in</strong> event-level semantics; it<br />

is a source applicative because <strong>the</strong> raised possessor of <strong>the</strong> r<strong>in</strong>g is <strong>the</strong> source ra<strong>the</strong>r <strong>the</strong> goal of <strong>the</strong><br />

transfer of possession. Interest<strong>in</strong>gly enough, <strong>the</strong> typically <strong>in</strong>alienable bare noun <strong>in</strong> this<br />

construction turns out to be fairly abstract <strong>in</strong> mean<strong>in</strong>g as we might expect of an <strong>in</strong>alienable noun<br />

that is non-specific. The similarities to body-part nouns that function as adpositions seems fairly<br />

clear.<br />

7


In (8), igi normally means ‘eye’ but <strong>in</strong> <strong>the</strong> context of <strong>the</strong>se constructions it comes to mean<br />

‘perception’ <strong>and</strong> it is <strong>the</strong> perception of <strong>the</strong> raised possessor that is transferred from <strong>the</strong> perceived<br />

entity to <strong>the</strong> ergative agent of <strong>the</strong> clause. Now <strong>in</strong> <strong>the</strong> Old Babylonian corpus that served as <strong>the</strong><br />

basis for my study, <strong>the</strong>re are only sixteen complex predicates that occur <strong>in</strong> <strong>the</strong> particular k<strong>in</strong>d of<br />

possessor rais<strong>in</strong>g construction <strong>in</strong> examples (8) through (11) <strong>and</strong> also <strong>in</strong>clude an <strong>in</strong>alienable noun<br />

as <strong>the</strong> direct object of <strong>the</strong> verb; <strong>the</strong>se are listed <strong>in</strong> (14). Of <strong>the</strong>se sixteen verbal complexes, five<br />

are verbs of direct perception, while seven seem to <strong>in</strong>volve some form of adversity or deprivation<br />

with respect to <strong>the</strong> raised possessor. No verbs of <strong>in</strong>direct perception of <strong>the</strong> ‘look at’ variety occur<br />

<strong>in</strong> <strong>the</strong> set.<br />

While Pylkkänen’s typology of <strong>applicatives</strong> <strong>and</strong> her transfer of possession model are truly<br />

<strong>in</strong>sightful, I would like to re-exam<strong>in</strong>e <strong>the</strong> syntactic structure that Pylkkänen proposes. Pylkkänen<br />

treats <strong>the</strong> applicative head, <strong>the</strong> *bi- prefix <strong>in</strong> <strong>Sumerian</strong>, as a higher order predicate that takes <strong>the</strong><br />

verbal lexeme, <strong>the</strong> direct object <strong>and</strong> <strong>the</strong> raised possessor as arguments, presumably <strong>in</strong> some k<strong>in</strong>d<br />

of VP shell as <strong>in</strong> (15). The possessive/dative phrase, namely gig ‘wheat’, raises <strong>in</strong>to <strong>the</strong><br />

postpositional phrase immediately above it <strong>and</strong> <strong>the</strong> postpositional phrase as a whole <strong>the</strong>n moves<br />

entirely out of <strong>the</strong> VP. The ergative DP was probably never <strong>in</strong> <strong>the</strong> VP to start with, so we are left<br />

with <strong>the</strong> surface order of constituents <strong>in</strong> (15) with <strong>the</strong> exception of <strong>the</strong> rais<strong>in</strong>g of <strong>the</strong> possessor<br />

out of <strong>the</strong> VP. The problem with (15) is that it seems to obscure <strong>the</strong> <strong>the</strong>matic relation between<br />

<strong>the</strong> possessor <strong>and</strong> <strong>the</strong> possessed entity, so tak<strong>in</strong>g <strong>the</strong> possessor rais<strong>in</strong>g construction as a model,<br />

we might end up <strong>in</strong>stead with someth<strong>in</strong>g like (17): with <strong>the</strong> <strong>in</strong>troduction of <strong>the</strong> low applicative<br />

head <strong>in</strong> (17b), igi ‘eye’ starts out as <strong>the</strong> complement of <strong>the</strong> low applicative head with its<br />

possessor coded <strong>in</strong> a dative or postpositional phrase. gig ‘wheat’ moves <strong>in</strong>to <strong>the</strong> specifier<br />

8


position of <strong>the</strong> dative postpositional phrase <strong>and</strong> <strong>the</strong>n <strong>the</strong> entire postpositional phrase moves out<br />

of <strong>the</strong> VP. In (17c), igi moves <strong>in</strong>to <strong>the</strong> specifier position of <strong>the</strong> low applicative <strong>and</strong> <strong>the</strong> low<br />

applicative phrase as a whole moves <strong>in</strong>to <strong>the</strong> specifier of VP. If noth<strong>in</strong>g else, <strong>the</strong> scenario <strong>in</strong> (17)<br />

moves igi through <strong>the</strong> <strong>the</strong>matic positions that it needs to move through <strong>and</strong> also prevents any<br />

fur<strong>the</strong>r modification of <strong>the</strong> noun that forms a component of <strong>the</strong> predicate. Now, even though <strong>the</strong><br />

nom<strong>in</strong>al component of this complex is buried deep under several layers of syntactic material <strong>and</strong><br />

should presumably have noth<strong>in</strong>g fur<strong>the</strong>r to contribute to <strong>the</strong> derivation of <strong>the</strong> clause, I would like<br />

to suggest that <strong>the</strong> nom<strong>in</strong>al component is still affected by specificity contrasts as def<strong>in</strong>ed by<br />

Dies<strong>in</strong>g’s <strong>mapp<strong>in</strong>g</strong> hypo<strong>the</strong>sis.<br />

[[The <strong>mapp<strong>in</strong>g</strong> hypo<strong>the</strong>sis]]<br />

Dies<strong>in</strong>g’s <strong>mapp<strong>in</strong>g</strong> hypo<strong>the</strong>sis states that DPs that rema<strong>in</strong> with<strong>in</strong> <strong>the</strong> VP are accorded a k<strong>in</strong>d of<br />

default existential quantification that leaves <strong>the</strong>m, or perhaps better, def<strong>in</strong>es <strong>the</strong>m as non-<br />

specific, while DPs that move out of <strong>the</strong> VP can act as <strong>the</strong> restriction for o<strong>the</strong>r quantifiers <strong>in</strong> <strong>the</strong><br />

clause <strong>and</strong> can <strong>the</strong>refore be described as specific. The important th<strong>in</strong>g is that any particular DP<br />

can only act as <strong>the</strong> restriction for one particular quantifier. That is to say, a nom<strong>in</strong>al phrase<br />

cannot both undergo existential closure with<strong>in</strong> <strong>the</strong> VP <strong>and</strong> <strong>the</strong>n later also act as <strong>the</strong> restriction for<br />

ano<strong>the</strong>r quantifier <strong>in</strong> <strong>the</strong> clause. With <strong>the</strong> cont<strong>in</strong>u<strong>in</strong>g development of <strong>the</strong> notion of semantic<br />

<strong>in</strong>corporation or more generally <strong>the</strong> view that non-specific nom<strong>in</strong>al phrases denote properties<br />

ra<strong>the</strong>r than <strong>in</strong>dividuals, <strong>the</strong> hypo<strong>the</strong>sis can also be turned on its head so as to argue that if a noun<br />

needs to be non-specific so as to restrict its predicate ra<strong>the</strong>r than referr<strong>in</strong>g to some entity, <strong>the</strong>n it<br />

should rema<strong>in</strong> with<strong>in</strong> <strong>the</strong> VP.<br />

9


In <strong>the</strong> contrastive focus construction <strong>in</strong> (18), <strong>the</strong> topic ama-bi is a def<strong>in</strong>ite DP <strong>and</strong> acts as <strong>the</strong><br />

quantifier for <strong>the</strong> clause as a whole, <strong>the</strong> focus phrase tilla iri-za-ka-åe nu-am ‘it is not to <strong>the</strong><br />

square of your city’ is <strong>the</strong> nuclear scope <strong>in</strong> <strong>the</strong> clause <strong>and</strong> <strong>the</strong> only rema<strong>in</strong><strong>in</strong>g part of sentence,<br />

namely bibsared—mean<strong>in</strong>g ‘that <strong>the</strong>y should be made to run’—must be mapped <strong>in</strong>to a<br />

restriction on <strong>the</strong> quantifier ama-bi. The example <strong>in</strong> (18) <strong>the</strong>refore yields someth<strong>in</strong>g like: “As for<br />

<strong>the</strong> mo<strong>the</strong>rs, it is not to <strong>the</strong> square of your [= Gilgamesh’s] city that <strong>the</strong>y should be made to run.”<br />

Now, translat<strong>in</strong>g <strong>the</strong> topic/focus articulation <strong>in</strong>to <strong>the</strong> quantificational language that Dies<strong>in</strong>g uses,<br />

<strong>the</strong> topic, ama-bi, is <strong>the</strong> quantifier; <strong>the</strong> restriction on <strong>the</strong> quantifier is <strong>the</strong> restrictive relative at <strong>the</strong><br />

far right of (18), <strong>and</strong> <strong>the</strong> nuclear scope—<strong>the</strong> part of <strong>the</strong> sentence that is be<strong>in</strong>g asserted ra<strong>the</strong>r than<br />

presupposed—corresponds to <strong>the</strong> focus phrase that is marked by <strong>the</strong> copula <strong>and</strong> immediately<br />

precedes <strong>the</strong> verb. S<strong>in</strong>ce <strong>the</strong> verb bibsared ‘that <strong>the</strong>y should be made to run’ acts as <strong>the</strong><br />

restriction on <strong>the</strong> topicalized DP ama-bi, it must move out of <strong>the</strong> VP so as to avoid existential<br />

closure. I would suggest that <strong>the</strong> underly<strong>in</strong>g head of what seems to be an <strong>in</strong>ternally headed<br />

restrictive relative, namely ama, pied-pipes <strong>the</strong> relative <strong>in</strong>to IP as predicted by Basilico’s study<br />

of head placement <strong>in</strong> <strong>in</strong>ternally headed relatives (1996) <strong>and</strong> is subsequently topicalized out of<br />

<strong>the</strong> relative <strong>and</strong> <strong>in</strong>to clause-<strong>in</strong>itial position. The important distributional fact is that low<br />

<strong>applicatives</strong>, namely <strong>the</strong> verbal complexes that select for a non-specific nom<strong>in</strong>al component as <strong>in</strong><br />

(8) do not occur <strong>in</strong> <strong>the</strong> type of contrastive focus construction <strong>in</strong> (18) apparently because <strong>the</strong> non-<br />

specific bare noun must rema<strong>in</strong> with<strong>in</strong> <strong>the</strong> VP if it is to restrict <strong>the</strong> predicate.<br />

[[Conclusion]]<br />

Thus, as long as <strong>the</strong> verbal complex rema<strong>in</strong>s with<strong>in</strong> <strong>the</strong> VP, <strong>the</strong> nom<strong>in</strong>al component is non-<br />

specific, denotes a property, <strong>and</strong> forms part of <strong>the</strong> predicate as part of a low applicative. When,<br />

10


however, <strong>the</strong> verbal complex moves out of <strong>the</strong> VP, several th<strong>in</strong>gs happen: <strong>the</strong> nom<strong>in</strong>al<br />

component becomes referential <strong>and</strong> specific <strong>and</strong> <strong>in</strong> (18), for example, can be topicalized, or<br />

alternatively, it can appear under contrastive focus as <strong>in</strong> (19). In both cases, <strong>the</strong> verb is<br />

<strong>in</strong>terpreted as causative ra<strong>the</strong>r than applicative.<br />

11


<strong>Low</strong> <strong>applicatives</strong> <strong>and</strong> <strong>the</strong> <strong>mapp<strong>in</strong>g</strong> hypo<strong>the</strong>sis <strong>in</strong> <strong>Sumerian</strong><br />

J. <strong>Cale</strong> Johnson (cale@ucla.edu)<br />

Cuneiform Digital Library Initiative, UCLA<br />

(1) NP-Rais<strong>in</strong>g as a form of possessor “rais<strong>in</strong>g” <strong>in</strong> English<br />

[ NP two pictures [ XP of [ DP John’s [ NP two pictures]]]]<br />

(2) English existential sentence with raised NP <strong>in</strong> def<strong>in</strong>iteness effect position<br />

There are [ NP two pictures of John’s] <strong>in</strong> <strong>the</strong> attic.<br />

(3) Possessor rais<strong>in</strong>g <strong>in</strong> Biblical Hebrew (Waltke <strong>and</strong> O’Connor 1990, 157)<br />

rΩ’îtî ben le-yiåΩ y<br />

I.saw a.son Dat-Jesse<br />

I saw a son of Jesse’s<br />

(4) Possessor Rais<strong>in</strong>g <strong>in</strong> <strong>Sumerian</strong> (Zólyomi 1999, 231, ex. 25)<br />

[N<strong>in</strong>girsu-ra]i aåa ki’aÑa-ni åu-[ni]i-a mu-ni-n-gi<br />

PN-Dat.Anim field beloved-3Poss h<strong>and</strong>-3Poss-Loc Vent-NI-Agr-return<br />

He (= Eannatum) returned (lit. returned to <strong>the</strong> h<strong>and</strong> [åu—gi]) his beloved field to<br />

N<strong>in</strong>girsu<br />

(5) Possessor rais<strong>in</strong>g with unspecified/<strong>in</strong>animate recipient (Zólyomi 1999, 231, ex. 27)<br />

Eniggal nub<strong>and</strong>a-e e-a åu-a bi-n-gi<br />

PN overseer-Erg house-Loc h<strong>and</strong>-Loc BI-Agr-return<br />

Eniggal, <strong>the</strong> overseer, returned <strong>the</strong>m (= various sorts of wood) <strong>in</strong> <strong>the</strong> house<br />

(6) Ord<strong>in</strong>ary use of dative as recipient with dative verbal prefix (VS 14, 27)<br />

Eniggal-e Lugalgaba gudu Baba-k-ra e-na-n-sum<br />

PN-Erg PN priest PN-Gen-Dat.Anim E-Dat.Anim-Agr-give<br />

Eniggal gave (a number of pots) to Lugalgaba, <strong>the</strong> gudu-priest of Baba<br />

(7) Dative prefix, NI series, possessive pronoun <strong>and</strong> demonstrative<br />

Dat NI Possessive pronoun Demonstrative<br />

1º ºa- ay- -Ñu N/A<br />

2º ra- ri- -zu N/A<br />

3ºAnim na- ni- -(a)ni ne(n)<br />

3ºInan ba- bi- -bi be(n)<br />

(8) <strong>Low</strong> source applicative with <strong>in</strong>animate raised possessor<br />

(Enmerkar <strong>and</strong> <strong>the</strong> lord of Aratta, l<strong>in</strong>e 554)<br />

en aratta-k-e gig-e igi bi-n-du<br />

lord GN-Gen-Erg wheat-Dat.Inan eye Ap/Caus.Inan-Agr-?<br />

The lord of Aratta saw <strong>the</strong> wheat<br />

12


(9) <strong>Low</strong> source applicative with animate raised possessor<br />

(Gudea Cyl. A, col. i, l<strong>in</strong>es 17-18)<br />

maåÑek-a gudea-Ø n<strong>in</strong>Ñirsuk-ra igi mu-ni-n-du-am<br />

night.vision-Loc PN-N/A PN-Dat.Anim eye Vent-Ap/Caus.Anim-Agr-?-Cop<br />

Gudea saw N<strong>in</strong>girsu <strong>in</strong> a night-vision<br />

(10) Causative construction with <strong>in</strong>animate causee<br />

(Nanna-Suen’s journey to Nippur, l<strong>in</strong>e 271; Zólyomi 1999, 218, ex. 3)<br />

ud-e niÑara-Ø bi-b-gu-en<br />

goat-Dat.Inan groats-N/A Ap/Caus.Inan-Agr-eat-1Sg.N/A<br />

I am <strong>the</strong> one who made <strong>the</strong> goats eat groats<br />

(11) Causative construction with animate causee<br />

(Idd<strong>in</strong>dagan Hymn A, l<strong>in</strong>e 160; Zólyomi 1999, 217, ex. 2)<br />

nugig-ra ki sikil-a n<strong>in</strong>da-Ø mu-ni-gu-e<br />

priestess-Dat.Anim place pure-Loc bread-N/A Vent-Ap/Caus.Anim-eat-Dur<br />

He is mak<strong>in</strong>g <strong>the</strong> priestess eat bread <strong>in</strong> <strong>the</strong> pure place<br />

(12) Korean transitive verb with possessed direct object<br />

totwuk-i Mary-ui panci-lul humchi-ess-ta<br />

thief-Nom Mary-Gen r<strong>in</strong>g-Acc steal-Past-Pla<strong>in</strong><br />

The thief stole Mary’s r<strong>in</strong>g (with no implication as to <strong>the</strong> prsence of Mary at <strong>the</strong> <strong>the</strong>ft)<br />

(13) Korean low source applicative (Pylkkänen 2002, 21)<br />

totwuk-i Mary-han<strong>the</strong>y panci-lul humchi-ess-ta<br />

thief-Nom Mary-Dat r<strong>in</strong>g-Acc steal-Past-Pla<strong>in</strong><br />

The thief stole a r<strong>in</strong>g from Mary (<strong>and</strong> it was <strong>in</strong> her possession when he stole it)<br />

(14) <strong>Low</strong> source applicative predicates <strong>in</strong> <strong>Sumerian</strong><br />

Noun Verb Mean<strong>in</strong>g Type<br />

a 2 dar ‘to confiscate’ Adversity<br />

a 2 Ñar ‘to lose’ Adversity<br />

a 2 su 3 ‘to spread out’ ?<br />

du 10 Ñal2 ‘to fight’ Adversity<br />

dur 2 Ñar ‘to sit; to seat’ ?<br />

Ñiå tuk ‘to hear’ Perception<br />

gu 2 du 3 ‘to neglect’ Adversity<br />

gu 2 e 3 ‘to wear, to cover’ ?<br />

gu 2 gur ‘to bend down; to spy’ Perception<br />

igi du 8 ‘to see’ Perception<br />

igi kar 2 ‘to exam<strong>in</strong>e’ Perception<br />

ka keå2 ‘to tie up; to b<strong>in</strong>d’ Adversity<br />

åa3 dab 5 ‘to be angry; to be worried’ ?<br />

umb<strong>in</strong> la 2 ‘to scratch’ Adversity<br />

za 3 tag ‘to push away; to reject’ Adversity<br />

zu 2 gub ‘to taste’ Perception<br />

13


(15) [ VP igi [ Vº bi [ VP du [ PP -e [ DP gig …]]]]]<br />

[ VP eye [ Vº LoAp [ VP hold/spread [ Poss/DatP Poss/Dat [ DP wheat …]]]]]<br />

(16) [ NP two pictures [ XP of [ DP John’s [ NP two pictures]]]]<br />

(17) a. [ VP du [ DP gig [ NP igi]]] (Genitive)<br />

b. [ VP du [ LoAp bi [ NP igi [ PP Dat [ DP gig] (Possessor rais<strong>in</strong>g)<br />

c. [ VP [ LoAp [ NP igi] bi] du …] (Dative moves out of VP, NP<br />

<strong>and</strong> LoAp roll up)<br />

(18) Negative contrastive focus construction (Gilgamesh <strong>and</strong> Huwawa, l<strong>in</strong>e 84)<br />

[TopP ama-bii [FocP tilla iri-zu-ak-åe nu-am<br />

[ TopP mo<strong>the</strong>r-3Pos i [ FocP square city-2Poss-Gen-Ter Neg-Cop<br />

[IP bi-b-sar-ed [VP e …]]]]<br />

[ IP Ap/Caus-Agr-run-Mod [ VP e …]]]]<br />

[TopP As for <strong>the</strong>ir mo<strong>the</strong>rs<br />

[FocP it is not TO THE MARKETPLACE OF YOUR CITY<br />

[IP that <strong>the</strong>y should be made to run ...<br />

Context: (Enkidu is speak<strong>in</strong>g to Gilgamesh) “Gilgamesh, for how long will you sleep?<br />

The sons of your city who came with you should not have to wait at <strong>the</strong> foot of <strong>the</strong> hills.<br />

As for <strong>the</strong>ir mo<strong>the</strong>rs, it is not to <strong>the</strong> marketplace of your city that <strong>the</strong>y should be made to<br />

run” (but ra<strong>the</strong>r it is to <strong>the</strong>ir sons that should run).<br />

(19) Complex predicate <strong>in</strong> negative contrastive focus construction (Lugalb<strong>and</strong>a II, l<strong>in</strong>e 215)<br />

[ TopP gu 3 i.ri.de 2.a nam mu.ni.ib.be 2.en … åeå-zu-ne-r<br />

[ TopP what I have told you <strong>the</strong> fate I have fixed for you] … bro<strong>the</strong>r-2Poss-Pl-Dat.Anim<br />

[FocP pa nu-am [ IP bi-b-e-en [ VP ...]]]]<br />

[FocP branch/w<strong>in</strong>g Neg-Cop [ IP Ap/Caus.Inan-Agr-go.out-1Sg.N/A [ VP<br />

[ TopP As for what I have told you, as for <strong>the</strong> fate I have fixed for you<br />

[ FocP it is not THE BRANCH/WING<br />

[ IP that you should cause to go forth to your bro<strong>the</strong>rs<br />

Context: (Anzu speak<strong>in</strong>g to Lugalb<strong>and</strong>a:) “Come now, my Lugalb<strong>and</strong>a. I shall give you<br />

some advice. May my advice be heeded. I shall say words to you. Bear <strong>the</strong>m <strong>in</strong> m<strong>in</strong>d!<br />

As for what I have told you, <strong>the</strong> fate I have fixed for you, do not make it known to your<br />

bro<strong>the</strong>rs.”<br />

1Sg.N/A = first person s<strong>in</strong>gular, nom<strong>in</strong>ative/absolutive<br />

2Poss = second person possessive pronoun<br />

3Poss = third person possessive pronoun<br />

Agr = agreement<br />

Ap/Caus.Anim = animate applicative/causative<br />

14


Ap/Caus.Inan = <strong>in</strong>animate applicative/causative<br />

Cop = copula<br />

Dat.Anim = animate dative<br />

Dat.Inan = <strong>in</strong>animate dative<br />

Dur = durative<br />

Erg = ergative<br />

Gen = genitive<br />

GN = geographical name<br />

Loc = locative<br />

Mod = future modal<br />

N/A = nom<strong>in</strong>ative/accusative<br />

Neg = negation<br />

Pl = plural<br />

Ter = term<strong>in</strong>ative (allative) case<br />

Vent = ventive, <strong>in</strong>dicat<strong>in</strong>g motion away from speaker<br />

Bibliography<br />

Basilico, David. 1996. Head position <strong>and</strong> <strong>in</strong>ternally headed relatives clauses. Language 72:<br />

498-532.<br />

Chung, S<strong>and</strong>ra, <strong>and</strong> William A. Ladusaw. 2004. Restriction <strong>and</strong> Saturation. L<strong>in</strong>guistic Inquiry<br />

Monographs 42. Cambridge: MIT Press.<br />

Dies<strong>in</strong>g, Molly. 1992. Indef<strong>in</strong>ities. L<strong>in</strong>guistic Inquiry Monographs 20. Cambridge: MIT Press.<br />

Guasti, Maria Teresa. 1993. Causative <strong>and</strong> Perception Verbs: A Comparative Study. Rosenberg<br />

<strong>and</strong> Sellier.<br />

Hallman, Peter. 2004. NP-<strong>in</strong>terpretation <strong>and</strong> <strong>the</strong> structure of predicates. Language 80(4):<br />

707-775.<br />

Harley, Heidi. 2003. Possession <strong>and</strong> <strong>the</strong> double object construction. Yearbook of L<strong>in</strong>guistic<br />

Variation 2: 29-68.<br />

Johnson, J. <strong>Cale</strong>. 2004. In <strong>the</strong> Eye of <strong>the</strong> Beholder: Quantificational, pragmatic <strong>and</strong> aspectual<br />

features of <strong>the</strong> *bi- verbal prefix <strong>in</strong> <strong>Sumerian</strong>. PhD dissertation, UCLA.<br />

––––––. forthcom<strong>in</strong>g. Internally Headed Relative Clauses <strong>in</strong> Akkadian: Identify<strong>in</strong>g Weak<br />

Quantification <strong>in</strong> <strong>the</strong> Construct State. To appear <strong>in</strong> Journal of Cuneiform Studies.<br />

Kayne, Richard. 1975. French Syntax: The Transformational Cycle. MIT Press.<br />

––––––. 1994. The Antisymmetry of Syntax. Cambridge: MIT Press.<br />

L<strong>and</strong>au, Idan. 1999. Possessor rais<strong>in</strong>g <strong>and</strong> <strong>the</strong> structure of <strong>the</strong> VP. L<strong>in</strong>gua 107: 1-37.<br />

Larson, Richard. 1988. On <strong>the</strong> Double Object Construction. L<strong>in</strong>guistic Inquiry 19(3):<br />

Moro, Andrea. 1997. The Rais<strong>in</strong>g of Predicates: Predicative Noun Phrases <strong>and</strong> <strong>the</strong> Theory of<br />

Clause Structure. Cambridge University Press.<br />

Pylkkänen, Marili<strong>in</strong>a. 2000. Deriv<strong>in</strong>g Adversity. WCCFL 19, ed. Roger Billerey, et al.<br />

Somerville, MA: Cascadilla Press.<br />

––––––. 2002. Introduc<strong>in</strong>g Arguments. PhD dissertation, MIT.<br />

Van Geenhoven, Veerle, <strong>and</strong> Louise McNally. 2005. On <strong>the</strong> property analysis of opaque<br />

complements. L<strong>in</strong>gua 115(6): 885-914.<br />

Zólyomi, Gábor. 1999. Directive <strong>in</strong>fix <strong>and</strong> oblique object <strong>in</strong> <strong>Sumerian</strong>: An account of <strong>the</strong> history<br />

of <strong>the</strong>ir relationship. Orientalia 68: 215-253.<br />

15

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!