Haviland - Fighting With Words

7/25/2019 Haviland - Fighting With Words

http://slidepdf.com/reader/full/haviland-fighting-with-words 1/9

Fighting Words: Evidential Particles, Affect and Argument

John B. Haviland

Reed College

We ethnographers, like everyone else, normally meet words n the quotidian

contexts of their daily lives, where, at least in principle, we should be able to

overcome our idealizations to see the richness, rather than the poverty,

of

language

strUcture. We find that languages not only permit the expression of, but also

grammaticalize aspects

of

daily life that we have not often

een

trained to detect.

Recently I have been looking at fights, in which people war with each othe r with

words as their weapons. The linguistic facts in verbal battles seem to collapse

or

conflate referential, expressive, and other rhetorical speech functions. The material

has led me to examine the range

of

linguistic devices which typically carry affective

and argumentative load, or which seem peculiarly suited to verbal battles. Here, of

course, linguistic device must be understood to include everything from

emotively charged lexical items to intonation, from anaphora and ellipsis to

gestures, and from poetic parallelism to particles. This paper is about evidential

particles, especially n Tzotzil and Guugu Yirnidhirr argument

Let me start, though, with English. We often fight with truth and the basic

techniques

of

contentiousness are often inseparable from the same matters that are

routinely encoded in the grammatical category of evidence: truth, reliability,

knowledge, and authority--relative to the context of the speech event. This is,

among other things, what irony is all about. Consider the two fragments in (I) and

2), where I have put some notionally evidential elements into boldface.

(1) The

Bickersons

(an

old-time radio

show

with Frances

Langford

and Don Ameche) :

Blanche; had a miserable t ime.

i t was the UNhappiest anniversary I ever spent.

Why

didn t you

show up for

the party, John7=

John; =1 TOLD

ya

I got stuck a t the office.

Blanche;

I d

like to believe that.

What

were you

doing?

John; working.

Blanche; su:re sure.

That s

always

the

f i rs t excuse.

(2)

Two

sis te rs

(aged

6

and

12)

fighting

s;

C'mon Maya,

STOP

i t .

m;

you

nearly

BROKE

the te levis ion=

s;

=

yeah

I

nearly

broke

the television.

Since evidentials grammaticalize aspects

of

the epistemological status of the

(putative) propositional substrate of utterances, they are by their very nature useful

in arming an argument over matters of fact. But there is usually more than this to

evidential particles: they are also interactive. Evidentials offer a delicate resource for

manipulating a constantly shifting common ground between speaker (in his or her

various faces) and interlocutors, a universe of discourse that has not only

epistemological but also moral character. Evidentials encode not only what a

Speaker knows or how he knows it; but also what an addressee can be taken to

know,

or

shouldknow, or apparently (perhaps culpabl y)fails to know. Again, this



is

what irony is often all about. For example, clause initial yu van, in Tzotzil

marks a proposition

as ridiculous or untenable, but at the same time presents it

somehow the alleged suggestion of some interlocutor, perhaps the present one.

As

in (3e)l, it typically elicits a demurring disclaimer.

(3) (d i s cus s ion o f t he

old days)

a .

1;

ti naka to 'ox to j t z -k ' e l

ART

j u s t

then p ine NONP+3E-watch

"They

only

used to use p i tch pine to s ee wi th . "

b . j ; naka

no'ox

j u s t only

"Tha t ' s

a l l . "

[

c . 1; i i

ta

a k ' u b a l t i k

ART ART

PREP night

"uh . . uh . . a t n igh t . "

d.

yu'van

oy Ius

un

EVID

ex i s t l igh t CL

"(Do you

suppose) they

had (e lec t r ic ) l igh t? "

e .

j ; ch 'aba l

to 'ox

not

then

"No

they

d idn ' t

(have l i gh t ) . "

.

~ x t e n d i n g

argument, the grammar of evidence picks out, presupposes,

or

Imphcates

VOIces

or faces (on both the speaker's

end

and that

of

his

interlocutors): those who do and don't,

or

can and can't, know. Kuroda (1973)

was among the fIrst to point out that grammar can accord special treatment to those

events or states, many

of

them psychological, which a t least in Japanese one can

only reliably predicate of oneself ('being sad,' for example). Grammatically, only

the experiencer

of

such states (or an imagined omniscient narrator)

is

entitled to use

what Kuroda calls a nonreportive description of such states and events, as in (4a).

(4)

(Kuroda

1973)

a .

Yamadera no

kane 0 ki i te , Mary wa kanas ika t t a

"Hear ing t he be l l

of

the

mountain

temple, Mary was sad."

/nonrepor t ive l

b . Yamadera

no

kane

0

k i i t e , Mary wa

kanas iga t ta .

"Hear ing t he be l l

of

t he mountain temple, Mary was sad."

I repor t ive with

g tl

By contrast, the gat form of (4b), appropriate to an evidentially less secure

report

of

someone else s state of mind, "has definite referential force directed toward the

'judger"'(p. 388). That is, the form "points semantically to the existence of a

subject of consciousness whose

judgment

the sentence is understood to

represent"(388), and who must be distinguished from the experiencer

of

the state

described. The outsider's lack of access to someone else's inner facts is here

morphologically encoded, and so, thereby, is his existence as a separate participant

indexed by the grammar.

Evidentials can also pick out or implicate those responsible for the issue of

truth, validity, or evidence in the first place. Consider how the participant structure

of a speech event is characteristically brought to the fore when evidentials appear

in

non-declarative sentences. There is a complicated, although by now familiar,

I

.;,

:

interaction between evidentials and illocutionary force. The connection between

dubitatives and interrogatives, for example, is iconically symbolized

by

frequently

shared morphology (if not

by

shared meaning, what Wierzbicka [1980] calls the

"ignorative"2). But there

is

more to this interaction. Both Warlpiri (Laughren 1981)

and Tzotzil have a hearsay particle (see [ aJ below) which marks the proposition of

a declarative as originating with,

or

vouched for by, someone other than the

speaker. Notably, the particle also appears in commands and questions, thus

nodding obliquely in the direction of otherwise unseen participants.

(5) War lp i r i (Laughren 1981)

nganta ' a f f i rma t ion from

ind i rec t

evidence,

hearsay '

a .

Marna- lu ma-nta

grass-PL

get-IMP

"Pick up the grass "

b .

Marna nganta-lu

ma-nta .

"They

say

you 've got to pick up the grass . "

(6) Tzotz i l Is 'hearsay,3

Mi l i ' - oxuk Is k ' a l a l i -O-ya l t an - e?

Q

here-2plA hearsay when

PAST-3A-fall ash-CL

"Were you

here when

t he

ashes f e l l

( implicates : somebody e l s e

wants

to

know)?"

As

a consequence, by indexing participants, evidentials drag us back again to the

arena where we should always have been: to situated speech and its unavoidably

social context.

Moreover, insofar as truth is something one (sometimes) predicates

of

propositions, whereas states of knowledge are properties of speakers and hearers,

evidentials bridge the treacherous and multi-tiered chasm between language and

metalanguage--a chasm we should by now fInd familiar, i no ~ e s s frightening.

I take as given an inherent multi unctionality (Silverstem 1985) to language,

so

that aspects

of

language design organized around certain linguistic functions, at

certain levels, may systematically feed other uses and purposes, at t h ~ r levels. A

single element (a demonstrative, for example, as part of an utterance) IS at once a

primary referential device (picking

out

a referent), a member

of

a structured

semantic domain (patterning both in sense and in syntax with other paradigmatically

similar elements), an indexical vehicle (tied inextricably to its moment and plaee,

and at the same time anchoring the utterance in precisely the ~ i g h t n:t0ment and

place), a functionally crucial part

of

the uttered token of an.lllocutlO.nary type

("identifying," perhaps, or simply "referring"), and element m practtcal soc al

act (so that its reduced pronominal character, say, Will contrast

With

an

a l t e r n ~ t l v e

way

of

"doing the same thing"--using a full noun I?hrase, for

~ x a m p l e

or a sIlent

gesture which would lend

to

the act of reference a different SOCIal character).

'Particles present the same sort

of

stratified

f u n t i ~ n ~ l

complexity, but in

spades. To bring this argument down to earth, let

e ~ l b i t

s o m e f r a g ~ e n t ~

?f

Tzotzil talk by way

of

introducing a few more eVldentlals. The Pcu:tlcies

In

question (here I will deal with only

half

a dozen

or

so

out of

an Inventory

considerably larger) fall into morpho-syntactic categories that suggest some of the

relevant complexities

of

scope and contrast .

There are "second position" clitics (Aissen, in press) which have restrIcted

distribution within a clause, and which are tied in scope to the corresponding



clausal predicate. The syntactic fa t C

particle/clitic) are rather complex h e ~

s ~ n c I ~ m ~ t h ~

precise placement

of

can be grouped together' They incl ' d u on ~ t n b u t i o n a l grounds the eVldel1tiallt

[the h ~ a y marker, w h i ~ h

we

hav u e most 1mportantly l 'they say, so I

canomcal examples. e already met], and nan 'perhaps's. (7)

7)

a.

b.

Mu la bu O-s-ve'

.

NEG HEARSAY

where

3A-3E-

"He d d n t eat .

ora

ea t

so it i s sa id .

nan O-s-botz '

a t

once

perha s 3A 3

lok 'e l

ta

' a n i l

"He probably ~ u l l e ~ ~ ? ~ ~ t l o u ~ klelaving

PREP f a s t

q u ~ c y .

.

T ~ e r e

are also clause-fina l cliti .

c o m t : ~ a t i o n s . The most notable exam I ~ s , , : ~ l c h may occur in various

by hi mdeed'), and

yu van

which pes are. a a nght, of course' (often preceded

'nonetheless' (examples in [S]). may also be glossed as 'of course' or perhaps

8) a. ja '

l i k

s -ve ' ta ora

MPH

a r i se

3E-eat

PREP h

(Tha t ' s r igh t )

h b

our

of

course

k'ox-on to ' ox

b ~ g a n to

eat

immediate ly ."

b.

small- lA then C L ' d d

I w ee

k

'oxas , ~ n d ~ e d

only

a

ch i ld then

-on

yu van .

.

small- lA

of

course

I

was smal l , of course "

The last word yu v I .o ' . an, a so occurs m ela .. . . .

a ye), where 1t means Do you suppo ? ~ s ~ l m . t i a l

p<;lsltlOn

(as we saw in [3d]

10 (?), it is often paired with the s ~ ~ : -- ut Implymg 'you would be wrong'

were It normally seems to beg rheto ' s ~ c o n d l?erson form of -na' 'know',

response. nca y or

an

1Oterlocutor's s e l f - d e f e n s i v ~

(9)

yu'van

ch ' abal

k

indeed

not exis t p r ~ x c h a n o chk'elvan

ana'oj

( erson

watch

You

don ' t

suppose fool ishl you

know

people

who ll y

that) the re are no

s ta re , (do

you?)"

~ i n a l l y there are evidential senten 'al '

look of It]' and ya el'it seems [by

th

ti lartieles, such as yilel 'it seems [by the

verbs

t

'see: and a i 'hear s ~ ~ ~ or feel of it].' These are derived from

r p p o S ~ I Y :

is transparently

d e r i ~ e d

.

f ~ r t h ~ r

evidential phrase, ta 'aiel

rom saymg: rom a say and means literally 'in saying,

(10)

a .

k ' e l - t z ' i ' yilel

see-dog

app 1

ch-O-bat

H

a ren t

y INC-3A-go

e

went to watch fo r

d '

b . ' a t a j j - ve ' - t ' k ' ogs ,

seems."

L

] -moton-t ik ya'el

C tha t

lE-eat -PL

lE -

ift

'

"Well

we a te it. t g - ~ L seems

, was a g ~ f t to us, it

seems."

c. Ja ' yech nolt i -O- ' ak' -b -a t

y-o l

ta 'alel-e

EMPH

thus

only COMP-3A-give-BEN-PASS

3E-child PREP saying.

"She was ju s t given an i l l eg i t ima te

ch i ld , supposedly ."

The etymology of these expressions suggests their kinship with a phenomenon,

noted by various authors

6

, linking evidential categories with explicit deictics and

perception verbs. The evidential element is directive: it points toward the relevant

evidence from which inferences may be drawn, and hence draws a contrast with an

unmarked proposition (which needs no special evidence). (lOa), thus, suggests: It

looked as

ifhe

was going (to the cornfield) to watch for dogs --suggesting what

sort

of

appearances were relevant to drawing this conclusion, and thereby priming

the hearer with the expectation that things were not as they seemed. (He was

actually heading for a lover's tryst.)1

In

the same vein, we discover, in the Trotzil phrase t 'aiel, an expression

which at last conforms to Jakobson's original characteriz.ation (1957) of the

evidential category: an indexical relation between the speech event, the narrated

event, and a narrated speech event (presumably, when someone told the speaker

about

the

narrated event). TheTzotz.il etymology directs attention

to

precisely such a

putative occasion of prior

speaking

8

•

In (l0e), the phrasing suggests She had an

illegitimate child, (or so she [or someone1 said).

Evidentials, indeed most particles, are notoriously resistant to uniform

analysis, in either propositional (semantic) or illocutionary terms

9

• Since such

particles do not pattern neatly into paradigmatic sets, they do not reward structural

treatment.

heir syntactic behavior is, as we see in the Tzotz.i1 case, heterogeneous,

and

it

presents daunting complexities

of

scope. To what, for example, does the

doubt of the clitic nan attach? When nan appears in standard second position

(following an introductory word

or

phrase and

any

temporal clitics) its scope seems

to extend to the entire clause

(lla);

but where it splits an idiomatic phrase (11

b) or

appears outside

of

second position its gaze settles, Janus-like, on constituents to

either side (see

[llc1

facing backwards to 'slingshot', and [lId] seemingly facing

forwards to 'girl').

11)

probable nan scope marked with

brackets

a . kultul to nan I i j ' a ' ye l e

a l ive still

probably

ART

person

"Probably

[ tha t

fe l low i s still a l ive ] .

b. te nan k ' a l a l mi j - k ' e l a

n

komel

the re probably when i f IE-give

away leaving

( te

k alal

means

'never mind,

forget

it )

I t probably

[doesn ' t

matter] if I ju s t give them

away."

c . muk'

bu

x-o - la j

ta

' u l i '

nan

s-bek'

y-a t ch -a ' i

NEG where AOR-3A-finish

PREP

slingshot.EVID 3E-seed

3E-penis

INc+3E-feel

"He f igu res

tha t probably

he

hasn ' t

been h i t in t he

ba l l s

[with

a s l ingsho t ] .

d. ' an soleI nan

tzeb

i - y - i k ' a ' a

yu 'van

PART only

EVID

g i r l

COMP-3E-mar

r

y EVID EVID

Why,

probably

he

married [a mere g i r l ] , of course , don ' t

you know."



As evidentials are clearly desi ed f '

characteristics are often relegated

t g ;l

or SItuated mteraction, many of

? t h e ~ interpersonal elements in l a ~ ~ ~ ~ : ( ~ t a b l e ~ ~ g m a t i c ~ s ~ d u u m along

lmphcatures, diminutives and au .

onon

ICS and sImtlar c O l l l v e : n t i o n l ~

psychological tinge.-indicating ~ ~ ~ ~ ~ t t v e s i l or ~ e r linguistic devices

w ~ e r e he stands with his f e l l o w ~ . - l e a v e ~ re:,

).

saYili

g

, h?w a speaker feels,

m t s ~ r a b l e ) . All the same, it is precisely ee eorehcally naked, wet,

busmess that the properties

of

Ian en spe get down to such

OlX1Uliint

Wittgenstein's phrase, "idling

lO

.

guage as a tool begIn

to

appear-·that it stops,

. I will limit myself to one famil f .

be

pnmary weapons in a war of wor l? uses. W.hy

e V I ~ n t i a l c ~ t e g o r i e s

should

have a look at some Tzotzil cases 'th s the

p a r h c u ~ a r

ISsue

of

Interest. So let's

Evidential s as we have'

n

a ew ~ m p a r a t i v e glances elsewhere.

between the

p r o p o ~ i t i o n a l

content of ' eXfrCItly grammaticalize a relationship .

b e l i ~ f ~ , attirudes, and intentions. Wh

an

u ~ r a n c e , and the speaker's knowledge

exphcItly at issue then as in certa' ere thIS

~ o w l e d g e

and these attitudes

provide a ~ e a n s 1 smuggling them i ~ n ~ ~ s thargument ove,r facts,

e ~ i d e n t i a l s

.

them, as It were, directly into words Th g e ~ r a m m a r ,

" " : I ~ O U t

haVIng to put .

appropriately crafted fonnulations'

n ~ t e

: r s g r ~ V l t d ~ h an. addItIonal resource for

contextually relevant knowled e

rath

a e Issue may be expected

M. Laughlin(l977:94) d e s c r i b ~ s er than ~ ~ O l u e , abstract knowledge

Robert

"accounts are sprinkled throughout i ~ ~ ~ : a

e,

~ m a c a n t . e c o myth-teller whose

t h ~ fonner a sign of his self-assured status i ; ~ ~ ~ S and ~ t u a l words a n ~ phrases;

pnde as a shaman, and an avowal of his inf ~ u r u t y , the latt;er a

s I ~ n

of his

he neglected to add the particle la which ' d ~ y ;lth the gods. QUIte dehberately .

he wants you to know that he was there : n t h ~ c : : ~ o ; ~ story

~ a S " O n 1 y

hearsay, for '

The ploy also works in reverse bal e creation.

Consider the stratagem of the skilled

,as

ver c o m b a ~ t s know (or soon learn).

boss, in (12). Called before a l a d i n } ( ~ ~ ~ ( ~ ~ ' m o u t h p l e ~ e for a village political

abuse of power in which a man who had -b

n

Ian)

~ u t h o n t y

to explain a blatant

pre ext returned to find his cornfield and

f

~ : : . e c a ed ~ " Y a y on a manufactured

facIle spokesman is given a chan

rul.

es sacnflced to a new road this

inteIJ?feter. It has already been

e s t ~ i 1 : ~ ~ ~ ~ h I S . ~ s s ' s h d e f e n s e

in Tzotzil to an

the VIllage when his lands were des e VICtim ad been lured away from

his account, subtly undennining the i : i t R ~ ; : b ~ ~ ~ : I Y inserts several las into '

(12) ~ a ~ g u m ~ n t a t damages hearing)

a,

Jun

J t a t a t i k l e ' 'une

ART one fa ther there CL

"That

o ld

gentleman

over the re "

b.

t a l

sk 'e jan

(sba) l i '

ta

l

.

co

k

~ s e n s y r o

'une

"Hmeh

neel

se l f

here

a t

lawyer CL

as Jcome, : beg

before

the

of f i c i a l s .

c. yu

un a Ja k'ux ta

yo'on

because CL EMPH

pain

in h is heart

"Because he

claims

to feel dis t ress "

d. komo muk' bu tey la 1 ' 1 ..

. vo

Je-e

~ ~ ~ c e

NEG

w?ere there CL ART yesterda -CL

s ~ n c e he c auns he

wasn' t

th y

ere yesterday."

349

I have already noted that such manipulated epistemological issues often leak

into other semantic areas, so that it should not surpriseus that evidentials also relate

to

questions

of

causation, volition, and agentivity (DeLancey 1985) a t the level

of

the

clausal encoding

of

events.

Still, face-to-face interaction, as the label implies, involves more than one

face. Doubt and hearsay may be individually expressed, but with agreement and

disagreement it takes two to tango. An important featureof evidential categories,

rarely mentioned in the literattIre on the subject, is their capacity to encode fearures

of what an interlocutor, as well as a speaker, knows or is ignorant of. Moreover,

such facts are no t absolute. The epistemological groundingof a conversation, the

presumed body

of

shared

information, is,

as

usual, a collaborative co-production

tailored to the purposes and conditions at hand. Since the extent of shared

knowledge between interlocutors can vary, it can also be a topic for contention,

or

for competitive interactional designs.

There are, however, some fonnal details. In Tzotzil, for example, the

evidential enclitics a a and yu van are logically tied to what conversation analysts

call "seconds"--tums that follow and are in some sense shaped by preceding rums.

Thus,

a a

means not only 'of course; or 'indeed' but more: 'I agree with that (and

what's more, I already knew it)' or, more contentiously, 'I can tell you that you're

right about what you've just said ' In fights, as in other fonns of verbal (not to

mention academic) exchange, it is often pressing business to establish precedence:

rights over and prior claims to infonnation.

Conversely, yu van, which I also glossed as 'indeed,' has a contradictory,

disagreeing tone. t means, 'indeed, despite what you have said (or implied)' and

goes on to suggest 'and you should have known it already '

In

some contexts the

particle seems to have the force

of

after all,'

as

in

'despite everything that has gone

before, it turns out after all thatp.

(A comparative digression: in Guugu Yimidhirr, two exclamations, yuu and

ngay, fulfill parallel functions: both respond to an interlocutor's remarks. The first

indicates that the speaker was just waiting for the other to come round to a truthor

proper fonnulation that he already possessed ("Yeah That's right "), The second

suggests, in the context

of

a question just asked, that the answer is somehow

coming out wrong or contrary to the speaker's expectations--a kind of surprised,

heckling, back-channel.)

Notably, an utterance with a a or

yu

van cannot stand alone. The (b)

sentences in (13) and (14) would not

be

well-fonned in

an

isolated frrst-tum.

12

(13) (conversa t ion

about

hybrid

corn)

a.

puta, 'unen k'ox-et ik

damn l i t t l e

small-PL

"Damn,

they ' r e just

l i t t l e

t iny

(kernels) "

b.

k'ox-et ik

a'a

"Yeah, they ' re

small ,

all

right "

(14)

( la ter in the same ta lk)

a. s - t a -o j kwentail

li

vojton ch-ak' 'uke

3E-get-STAT

account ART cob

INC+3E-give

a lso

I t gives a suff ic ient ly large cob a lso

(I

suppose?)"

ak ' -o

mi

k'ox-et ik yi le l

y-ok

give-IMP Q small-PL it appears 3E-stalk

"even if

i t s s ta lk appears small"



b. ch-ak' a'a yu'van

INC+3E-give CL CL

"Of coura. i t

does (c . i . :

how

could

you think

otherwise )"

Mary Laughren( 1981) notes

that

the Warl iri

..

. . .

f:; ':Ffoamad

u g g ~ ~ ~ the 'negati<.>n of a former p r e s u : p o s i J : ~ f 1 : ~ ? ~ i l ~ I ~ ~ ~ ~ ~ .

,an conu""",cts, somethmg that has been said

or

suggested before.

(15)

-Karlarra-lku

nganm-lu rdaku-ju pangurnu

w e s t ~ S E Q PP -3PL hole-DEL dug

-NgaYI. Kulanganm

yat i ja rra

::I've heard they dug the darn ~ u t west ."

Really. thought

i t

was north."

Similarly, in Guugu Yimidhirr a related P -N 1

utterance final clitics

ba

and

ga

both giossed again

: ; t n d ~ ~ ~ ~ ;nruage

the

on the matter of whether the speaker is a eein

or

disa in . n ~ preCisely

~ ~ ~ o ~ ~

~ ~ ~ ~ ~ ~ ~ j ~ : t been said, as tt l 6 ~ , but

o = m ~ s : : : p a ~ l ~ ~

(16) (talk

about

a countryman)

- j ; yubaal

guugu

nhanu-um-i yirrgaalgay?

" . 2DuNOM

word

2sGEN-CAT-LOC

were

speaking

D ~ d you two t a lk in

your language?"

- r ; nyulu guugu ngadhun-gal ngadhuu-m-ay yirrgaalgay

nyu

lu

-ugu ba.

3sNOM

wor d ls-ADES lsgGEN-CAT-LOC

3sNOM-EMPH

indeed

"He spoke to me in

my language,

yes he did "

(17) (myth ~ b o u t two feuding

kinsmen)

- ~ ~ nyulu

ngiinggirr

nhaadhi

gurra

3sNOM

snoring

saw

then

"(for a

long

time) . . he l istened for h

h

' t at

snoring."

- 0

n g ~ i n g g i r r

bulngaalngal ba

snoring

pulling indeed

nOh, so he

f inal ly

snoring "

(18)

(lost

countrymen unexpectedly return)

-dhanaan banydyi

3plACC waited

"He

waited

for them ( to corne up)."

-a bama yurra ga,

waarmbaadhi

man

2plNOM

indeed returned

' Ah, so

i t

S you al l You carne back. '"

-a waarmbaadhi nganhdhaan duday gurra

returned IplNOM ran th

.. 'y h en

es, we

ave come back. W ran away (from

there) ."

: : : r e : : ~ an: ~ C O r d with or < > n t r a ~ i c t not only verbal propositions.

that th

d

r an c ~ p a , ut also expectatIons:

In

(17) the confmned expectation

oth

e versary wlll fall asleep,

In

,(18) the d isconfmned presumption that the

ers would not return. The sense In whlch these particles can anticipate or invite a

positive

or

negative response is particularly clear in the stock Guugu Yimidhirr

evidential tag questions: yuu b "Isn't that soT'--which fishes for confirmation-

and

gaari ga

"No, that's not so "--bracing for further contradiction.

I have suggested th at evidentials are potent tools in verbal battle, as well as

in ordinary conversation, in part because they help negotiate common ground and

the universe

of

(moral) discourse. We know that some things (such as, say,

psychological states in Japanese) are by definition not part of common ground: they

are, in the unmarked case,

out

of bounds for shared or interpersonal scrutiny. In

this sense, evidentials help keep the fences in pl ace and in good repair, partitioning

the world

of

who is in a position to know, who has the right to know, who can

even claim to know about, the crucial facts

of

a situation. This brings me to my last

examples, from somewhat closer to home.

In fragment (19), from a deliciously violent

argument between

two

housemates (which end ed

in

their dissolving their agreement

to

share an apartment),

there is a striking use of evidential markers (applied with heavy sarcasm) to fan the

flames of argument. Notably, these Spanish speakers tum Japanese psychology on

its head: P denies L access to her own declared inner states. Note also the explicit

metalinguistic tactic, again with an evidential flavor: "I have been very worried

about you, and i you want me to tell you so, I'll tell you so,

(19)

(roommates squabble

in Mexico

City)

1; y

me

preocupe mucho por

ti

y

1;

p;

1;

p;

"and was very worried about you"

[

"yeah yeah yeah"

-10 diga

te

10

di:go:

[

y s i quieres que te -

"and i f

you want

me

to"

" te l l

you

so,

I ll t e l l you

so"

[

y

desde

que regrese

"and since I have come back"

[

he estado

muy preocupada por

ti

Pi lar

"I

have been very worried about

you"

p r e o c u p a d ~ i m a

mano

"very worried, friend"

[

SI pero

"yeah"

Finally, have a look at the whimpering evidentials of a disputed volleyball

serve, among a bunch of American academics. In an ambience of constant ironic

joking, and self-mocking put-downs, complaints, and criticism, when a real

disagreement emerges, the players must search for different rhetorical techniques.

Here the players take refuge in an increased dose of sincerity ("honestly"), coy

evidential framing ("saw t out"), token expressions of affect ("what a pity ), and

even explicit evidential meta-commentary ("I believe him"), to preserve their

civilized immunity from overt hostilities.



(20)

p;

n;

r ;

n;

db;

c ;

b;

r ;

p;

c ;

352

<academics

adjudicate a l ine

ca l l

in volleyball)I3

that

was good

bri l l iant

serve

No I honestly

didn ' t

think so

I thought

it

was out

I honestly thought it was out myself

what a pity

it s your cal l

what'd

you

think

Robert

it s your cal l '

I saw it out, Carol

huh

I believe him

I have tried to illustrate the multifunctionality

of

the ,

c e ~ n

evidential categories, in the sense that they act = ~ ~ expressl?D

discursive, and transactional levels simultaneou sly' more th th'

th

' pragmatic,

be a common m

ltif

al

an

IS ere seems to

; ~ i ~ ~ ~ n ~ j : ~ u

~ ~ r ~ ~ ~ i f ~ £ ~ ~ ~ ~ ~ ~ ~ ~ t i ? ~ ~

a ; ~ o ~ ~ ~ ~ : s ~

~ ~ ~ e i ~ ~ ;

implicit comment on moral d " ge, m . e c ~ n t e x t of speech; they

permit

prothbes

anthd

barbs in the m i n : r : : : f J ~ ~ : ~ ~ ~ ~ ~ ~ e ~ ~ ~ : o ~ p e r T h a e t e g a e n S einrtealractiral ve

IS at e nographic richn fid rd . mo

~ Y S i S , if we are to take ~ ~ ~ U ~ I ; ~ ; n c f ; f : ~ ~ ~ ~ ~ g ~ ~ ~ ~ ~ : s ~ n ~

~ ~ g ~ :

Notes

I All original I .

are taken from

t r a n s c J ~ : : r g o ~ e ~ a t i : ~ I l , Guugu

Yimidhirr, Spanish

and

English,

2 In Tzotzil the d b' tati i . .

th

bY

a separate f o ~

~ ~ n . ~ h ~ f o ; : e C o ? ~ o : : : , ~ : : : K ~ a ~ ~ : ~ ~ : ~ o ~ a t i : J d s e n t e n c e s ,

at from a

proposillon p

one can fo . e ressee, so

question-marker) which means Do

rm

a

questl°thn

Ml p van? (where

mi

is the

3 B ' . ' you Suppose

at

p?

lah which, ~ : r : t : : g ~ v : : : ~ ~ I J ; s ~ ~ I I s 7 ~ ~ e a , ~ i m i l a r o b s e r v ~ ~ o n about Tzeltal

on illocutionary force. ' ge not on proposItiOnal c ontent but

4 T h d . ,

e

secon

poSItIon" formulation as in rna I .

elaborate statement of what can

constitute'

he s'

lnr

a n g u a d ~ e s , requI eS an

Moreover some membe f th· I mg e prece mg constituent.

"second s i t i o n " clitics a ~ e

a ~ ~ ~ ~ ~ a l s ~ p p e a r

~ l s e w h e r e in

c l ~ u s e ,

Other

evidentials. Within each se . u empo meanIngs, and ordmanly precede

least in "second position." mantic class, the members

are mutually

exclusive,

at

( m i S t a k : ~ o m , e : h e r

infrequent mem >ers are kik 'I guess, maybe,'

ka

'I thought

than nan J ~ o u g t ~ ~ ~ m s to h ~ v e a shghtly more positive or optimistic tone to it

, sometimes occur together.

6 See for example Sil t (1978)

which p p e ~ t . ,vers , " on the Wasco passive of evidence. .

th . .

0

mcorporate an explICIt deiCtic

-ix

'there"

Hanks(l984)

describes

W ~ V I ~ ~ , t I a l

n a t l f ~

of

s t e n ~ i v e

deictics in Yucatec; and Laughren(l981) cites the

e v i d e ~ ~ , , ) X ~ s l t I ( ~ ~ , a l

Pf

hcle

" kari which indicates "supposition from direct

in

i s c o u ~

. un 'de

iS

vo ume) makes related observations

about

demonstratives

e, ll a WI range of languages. .

353

Len Talmy, in comments on the oral presentation of this talk, pointed out

that Russian vot, a presentational form meaning 'This (and here

it

is)' is difficult to

translate because

of

the situational vividness it conjures; it s usually restricted, even

in narrative. to present tense sentences. The phenomenon may be related to deictic

(directive evidential) force. Terry Kau fman suggests, in a simil ar spirit, the English

equivalent

10

7 Compare the evidential flavor (accompanied again by a presentational

vividness) in the colloquial English form of words

I

saw/see where p --which,

as

Len Talmy observes, does not easily admit a 2nd person subject (except in

questions. as in the

case of

Japanese psychological predicates), and which seems

otherwise pragmatically restricted.

8 The nature

of

quoted and reported speech is clearly

of

related interest.

Tzotzil uses the verb chi 'say' to bracket quoted or dramatized dialogue, and the

particle-like inflected form xi 'he says, one says' interacts with the hearsay clitic b

in a complicated way.

9 Nonetheless, there is persistent programmatic optimism in some semantic

circles. See Wierzbicka 1976, 1980, and Goddard 1979,

I

have

not

attempted to

provide semantically uniform and well-motivated formulas for the Tzotzil

evidentials described here, despite urgings from

Tim

Shopen that such an attempt is

necessary.

10 The confusions which occupy us arise when language is like an engine

idling, not when it is doing work"(1958 section 132).

11

DeLancey

(1986) shows

that

Tibetan

evidentials interact

with

interlocutors' assumptions about expected, predictable, contextually "normalized"

background knowledge,

a

phenomenon which he relates to the old/new

distinction.

12

Tzotzil speakers can articulate certain metapragmatic intuitions about these

particles;

I

can remember being criticized and mocked for misusing

a

a, both in

isolated fIrst-turns,

and in

situations where it was obvious that I could not know

enough about the topic at

hand to

be in a position to

agree

in the way that the

particle required.

In comments after the talk, a psychoanalyst in the audience pointed out his

own strategic, pragmatically ill-formed, use

of of

course as a provocative

and

deliberate prod to patients' frnmings of absolute certainty on some matter, which

could be challenged or cast into doubt by the analyst's

covert

suggestion that

he

too

was in possession

of

some

of

the relevant facts. .

Compare the pragmatic misfire involved with the misuse of the particle oh

(typically associated with news receipt, or, in a parallel way, with just remembering

something one was going to say) in a turn where deliberate and pre-planned matters

are mentioned.

13

These volleyball transcripts, and some of these thoughts, were collected

during

my

stay at the

Center

for Advanced Study in the Behavioral Sciences

at

Stanford in 1985-86. I

am

grateful for support from the Harry Frank Guggenheim

Memorial Foundation and NSF Grant #BNS-8011494, Fieldwork in Zinacantan

and Hopevale

has been supported by the Australian National University, the

Universidad

Nacional

Aut6noma

de Mexico,

and the Australian Institute of

Aboriginal Studies. I thank David French for helpful comments.



ibliography

Aissen, Judith in press. Tzotzil Clause Structure. Reidel.

Brown, Penelope and Stephen Levinson 1978. Universals in language usage:

politeness phenomena. In Esther N. Goody (00.), Questions and Politeness:

Strategies in Social Interaction , 56-289. Cambridge: Cambridge University

Press.

DeLancey, Scott 1985. Lhasa Tibetan evidentials and the semantics of causation.

Proc. of the Eleventh Annual Meeting of the Berkeley Linguistics Society

65-72.

DeLancey, Scott 1986. Evidentiality and volitionality in Tibetan. In Wallace Chafe

and Johana Nichols (OOs.), Evidentiality: The Linguistic Coding

of

Epistemology (Advances in Discourse Processes, vol. 20), pp. 203-213.

Norwood, N. .: Ablex.

Goddard, Cliff 1979. Particles and illocutionary semantics. Papers in Linguistics.

12:1-2:185-230.

Spring-Summer 1979.

Hanks, William F 1984. The evidential core of deixis in Yucatec Maya. n Joseph

Drogo, Veena Mishra and David Testen (eds.),

CLS 20: Papers from the

Twentieth Regional Meeting of he Chicago Linguist ics SoCiety 154-72.

Jakobson, Roman 1957. Shifters, verbal categories, and the Russian verb. Russian

Language Project, Dept.

of

Slavic Languages and Literatures, Harvard

University.

Kuroda, S-Y. 1973. In S, Anderson and Paul Kiparsky (oos.), A Festschriftfor

Morris Halle 377-391. New York: Holt, Rinehart and Winston.

Laughlin, Robert M. 1977,

f Cabbages

and

Kings: Tales from Zinacantan.

Washington, D,C.: Smithsonian Institution.

Laughren, Mary 1981. Propositional particles in Warlpiri. Paper presented to

Australian Linguistic Society meetings, Canberra, ACf. In Steven Swartz

(ed.), Papers in Warlpiri Grammar: In Memory of Lothar Jagst. S,I.L.,

Australia Branch.

Silverstein, Michae11978. Debds and deducibility

in

a Wasco-Wishram passive

of

evidence. Proc. of he Fourth Annual Meeting of he Berkeley Linguistics

Society.

Silverstein, Michael 1985.

The

functional stratification of language

and

ontogenesis. In James V. Wertsch (ed,),

Culture communication

and

cognition: Vygotskian perspectives pp. 205-235, Cambridge: Cambridge

University Press.

Wierzbicka, Anna 1976, Particles and linguistic relativity. International Review of

Slavic Linguistics. 1 2-3):327-367,

Wierzbicka, Anna 1980. Lingua Mentalis. Sydney: Academic Press.

X Bar alantics

Ray

Jackendoff

Brandeis Un;versity/Univ. of Arizona

I take the major concerns of semantic t ~ o r y to be

(1) the

form

of the mentally e n ~ o d ~ d

; n f o r m a t ~ o n t h ) a t

we

« t and (2) the pr1nc1ples usee ln a

call c ~ n c e ~ s, the basis of this information

and

perform1nQ 1nferences on f f

(b) relat1nQ

this

information

o t h ~ r

orms

0 t 1

information' used by

t h ~

human m1nd, 1 ~ C 1 U ~ i ~ ~ f ~ ~ m a ~ ~ 0 ~ (a

i ~ n ~ ~ ; : t ~ ~ 9 ~ ~ ) ) e ~ ~ ~ t ~ i t ~ ~ t ~ h s ~ : ~ ~ ~ ; : ~ ~ ~ ~ ~ ~ ~ ~ i t ~ ~ : s i a ~ ~ ~ ; ~ e s .

A semant1c theory mus e

formal components: , rules that collectively describe

(1) a set of formatlon f the lanQuaqe of

in finite form the e x p r e s s 1 ~ e power 0 f

thouQht para11elinQ, for 1nstance, the

set

0 'b1e

f o r m a t i ~ n

rules (the ~ r a m m a r ) that,de11neate POSS1

syntactic structures 1n lanquaqe, . . f te

(2)

a set

of

inference rules that descr1be

1n 1n1

form the a 1l0v/able der ivation s f r ~ m one

c o n c ~ P t u ~

1

expression

to

a n o t ~ e r ( t ~ e s e ~ a Y t . ' n ; ' l ~ ~ e ~ e ~ ~ ~ 0 1 0 Q i C a 1

invited inference

and

heur1S

1CS

,

entailments);

f . formation that conceptual

(3) for each other form 0 1n ,. f

information

can

be related

to,

a f1n1te s ~ t 0

correspondence rules that define the

~ a p p 1 n q .

. not

Under s u c ~ a ~ o n ~ e p t i o n , m u c h i ~ ~ e s ~ ~ : n ~ ~ ~ c ~ ~ ~ ~ { S

part

of

1 1 n g ~ ~ ~ t 1 ~ ~ c ~ e ~ h : e t h ~ O r y

is concerned are not

structures

Wl

d w

t

(Only

the correspondence rule

lanquage-depen en : . th lanQuaQe.) 1

component has spec1f1ca11

Y

l

dnoaryW'conservatism, that

e on

Qrounds

of

evo

u

10

.

assuf 1 ' . , - - bo t h hiqher animals

and

bab1es--

non11nQU1st1c

orQan1sms i th

r

mental

also p o ~ s e s s c o n c e P t ~ ~ ~ ~ ; r ~ ~ ~ ~ r : ~

o ~ r s , e ~ u t

formally

r e p ~ r t o 1 ~ e ,

perhaps

Th d fference

between

us

and

s

im1l

ar ,n

many

respects. e 1 , 1 d

the beasts is

that we

evolved a capac1ty to

e a r ~

~ ~ e

rocess

syntactic

and phonological structures an

~ a p i n Q S from them to conceptual structures

.and

to t ~ ~ us

auditor

and

motor peripheries. These mapp1nqs

perm

a r e 1 a t ~ v e l y overt

realization

of conceptual

structure

unavailable to o ~ h e r

organtiSms.

th t linQuists should not

However

th1S does no mean

a

'd s the

be

concerned'with semantic theory. LanguaQe provl e

1987 Ray Jackendoff



ERKELEY LINGUISTICS SOCIETY

Proceedings of the Thirteenth Annual Meeting

February 14-16 1987

GENER L SESSION

l ~ T I

P R SESSION

O

GR MM R ND

COGNITION

edited by

Jon Aske

N

atasha

Beery

Laura

Michaelis

Hana

Filip

Berkeley Linguistics

Society



MITHUN,

MARIANNE

.

The Grammatical Nature and

Discourse Power of Demonstratives 184

NORVIG, PETER and G EORGE LAKOFF

Taking: A Study in Lexical Network Theory ............................................... 195

OHALA, JOHN

Experimental Phonology ................................................................................... 2 7

PULLUM, GEOFFREY K.

see Zwicky and Pullum

RHODES, RICHARD

Paradigms

Large

and

Smal l .............................................................................

223

SCHILLER, ERIC

Which Way Did They Grow?

(Morphology and

the

Austro -Tai/( Macro )Austri c Deba te) ....................... 235

SHANNON, THOMAS F.

On

Some Recent Claims of Relational

Grammar

........................................ 247

SHAUL, DAVE

Pragmatic Constraints

on Hopi Narrative Discourse .................................. 263

SZATROWSKI, POLLY

A Discourse Analysis of Japanese Invita tions ............................................... 270

TAN,

FU

The Predicate

Argument

Structure of

Bei .................................................. 285

TUITE,

KEVIN

Ind irect Transitives in Geor gian ..................................................................... 296

VIJAYAKRISHNAN, K.G.

Hierarchical Representation of Phonological Features ............................... 310

WILSON, STEPHEN A.

The Deve lopment of Tones in Heiltsuk .......................................................... 321

ZWICKY, ARNOLD M.

and GEOFFREY

K. PULLUM

Plain

Morphology and Expressive Morphology ............................................ 330

R SESSION

HAVILAND, JOHN B.

Fighting Words: Evidential particles, Affect, and

Argument

.................... 343

JACKEt-. DOFF , RAY

X-Bar Seman tics ................................................................................................. 355

LAMBRECHT, KNUD

Aboutness as a Cognitive Category:

The

Thetic-Categori cal Distinction Revi sited .............................................. 366

LANGACKER. RONALD W.

Grammatical

Ramifications of

the

Setting/Participant Distinction ...... 383

LEE, MICHAEL

The Cognitive Basis of Classifier Systems ..................................................... 395

POTEET,

STEPHEN

Paths Through

Different Domains:

A Cognitive Grammar Analysis

of

Mandarin Dao .................................... 408

RICE, SALLY

Towards a Transitive Prototype:

Evidence from Some Atypical Engl ish Passives ........................................... 422

SLOBIN, DAN

Thi nki ng for Speaking ....................................................................................... 435

SWEETSER, EVE

Metaphorical Models of Thought and Speech:

A Comparison of Historical Directions

and

Metaphorical Mappings in

the

Two Domai ns ............................................... 446

WILKINS, WENDY

On the Linguistic Function

of Event

Roles ................................................... 460

* 4:

The

following papers were presented

at

the conference but do

not appear

in

this volume:

BATES, ELIZABETH

Function and Form in Language Acquisition

CLARK, EVE V.

Why

Comprehension Leads

Production

in Acquisition

KAY, MARTIN

Finite

State Morphology

MATSUMOTO, YO, ET AL.

The Japanese Reflexive Zibun-zishin

and the

Supposed Relatedness of Anaphora and

NP-trace

MORDECHAY, SUSAN

Pragmatic Scales

and

the Semantics of

Already

R l ; ~ t E L H A R T DAVID

Some Speculations on

the

Microstructure

of

Linguistic Info rmation Processing

THOMPSON, SANDRA

That-deletion in English from a Discourse Perspective

ZERNIK, URI

Acquiring Idioms from Examples in Context: Learning by

Explanation

The following papers were not presented at the conference but are included

in this volume:

POTEET, STEPHEN

Paths Through Different Domains:

A Cognitive

Grammar

Analysis

of Mandarin

Dao

VIJAYAKRISHNAN,

K.G.

Hierarchical Representatio n of Phonological Features

Documents

Haviland - Fighting With Words