Volume 2 Issue 1 (2003)
Note: Linguistic Discovery uses Unicode characters
to represent phonetic symbols. Please see Optimizing Display
for requirements to accurately reproduce this page.
A Particle of Indefiniteness in American Sign
We describe here the characteristics of a very frequently-occurring
ASL indefinite focus particle, which has not previously been recognized as such.
We show that, despite its similarity to the question sign
“WHAT”, the particle is distinct from that sign in terms of
articulation, function, and distribution. The particle serves to express
“uncertainty” in various ways, which can be formalized semantically
in terms of a domain-widening effect of the same sort as that proposed for
English ‘any’ by Kadmon & Landman (1993). Its function is to
widen the domain of possibilities under consideration from the typical to
include the non-typical as well, along a dimension appropriate in the
In this article, we describe a focus particle that occurs with great
frequency in American Sign Language (ASL) in indefinite contexts. This particle
has not been previously identified as such in the ASL literature. Although its
articulation is similar to that of the generic question sign conventionally
glossed as “WHAT”, the two signs are distinguishable in terms of
articulation, function, and distribution. Here we discuss the characteristics of
this particle and the kinds of contexts in which it occurs. It serves to express
“uncertainty” in a variety of ways. In the final section, we address
the semantics associated with this particle and propose that it has a
domain-widening effect comparable to that proposed by Kadmon and Landman (1993)
for any in English. This ASL particle functions to widen the domain of
possibilities under consideration along some contextually determined dimension.
2. Background about American
Despite the difference in modality, signed languages have the
same fundamental organization as spoken languages. Neidle et al. (2000) have
argued that the basic hierarchical structure of American Sign Language (ASL)
is as illustrated in Figure 1, and thus is completely comparable to what has
been proposed for spoken languages.
Figure 1. Skeletal structure of CP in ASL
2.1. Nonmanual Correlates of
One particularly interesting characteristic of signed languages is the
use of nonmanual expressions (gestures of the face and movements of the head and
upper body), which extend over phrasal domains, to express grammatical
information. This is one important use (of many) for nonmanual expressions in
Neidle et al. (2000:43-45) have proposed the following generalizations
for the distribution of nonmanual syntactic markings:
- Nonmanual syntactic markings are frequently associated with
syntactic features residing in the heads of functional
- The nonmanual marking may spread over the
c-command domain of the node with which it is associated (reflecting relations
- Spread of the nonmanual marking is
optional if manual material is available locally. However, in the absence of
such manual material, the marking spreads obligatorily so that it may be
articulated simultaneously with manual material.
intensity of the nonmanual marking is greatest at the node of origin and
decreases as distance from that node increases.
- As in
the manual channel, perseveration (maintenance of a particular articulation that
will recur) is found with nonmanual
Question constructions have been the subject of some controversy in the
literature, with respect to both the facts and
analysis. Neidle et al.
(2000:chapter 2) discuss possible contributing factors to the conflicting
reports in the literature, some related to methodological problems in eliciting
grammaticality judgments from native signers. The fact that most of the
constructions reported in the literature on ASL syntax are given only in gloss
form (using the nearest equivalent English translation for each ASL sign) makes
it difficult to evaluate claims and conclusions about ASL data. To avoid this
problem, the American Sign Language Linguistic Research Project (ASLLRP) has
made available digital video examples, signed by native signers of ASL,
illustrating the constructions we discuss in this paper and others. These can be
obtained on CD-ROM and at our Web site
For simplicity of exposition here, we will focus on those sentences that
have a single
The wh-phrase (e.g., WHO, WHERE, WHEN, etc.) may surface either in situ
or at the right periphery of the clause. There is a distinction between the two
positions in terms of interpretation: when the wh-phrase is at the right
periphery, it is necessarily interpreted as
focused. The distribution of the
nonmanual marking characteristic of wh-questions (labeled as ‘whq’
and consisting of a cluster of properties, including furrowed brows, squinted
eyes, and a slight side-to-side
differs. If the wh-phrase is in
situ, the wh-marking obligatorily spreads over the entire CP. In contrast,
when the wh-phrase occurs at the right periphery, spread of the marking over the
rest of the CP is not required. This provides the basis for an argument (Neidle
et al., 2000) that the final wh-phrase is in a right-peripheral Spec, CP: the
wh-phrase provides manual material local to the +wh feature of C, so spread of
the nonmanual +wh marking over the c-command domain of C is not obligatory.
Wh-constructions are illustrated in (1)–(8). Glossing conventions
for representing ASL sentences are explained in the Appendix. For further
details about conventions used for annotation, see Neidle (2002).
Hyperlinks: Example 1
Hyperlinks: Example 3
Wh-phrases at the right
Hyperlinks: Example 5
Hyperlinks: Example 6
Hyperlinks: Example 7
Hyperlinks: Example 8
There is something interesting about (7) as signed in the
video illustrating this sentence (in particular, with respect to the
non-dominant hand); this is not annotated here and we will return to discuss it
in section 3.5.
One of the common wh-signs is conventionally glossed as
“WHAT”. This sign involves a side-to-side shaking of the hands in
front of the chest, with a 5 handshape (all five fingers extended), palms facing
upward, as in the following sentences:
Hyperlinks: Example 9
Hyperlinks: Example 10
The sign glossed as “WHAT” can also occur in a
sentence-final tag with a construction that contains a different wh-sign, as
shown in (11).
Hyperlinks: Example 11
Discovery of a Previously Unidentified Particle
The construction illustrated in (12) and (13) involves a sign, glossed
here as ‘part:indef,’ that is similar to the final sign in
(9)–(11), but critically different from it.
Hyperlinks: Example 12
Hyperlinks: Example 13
In this construction, the final sign is articulated with the
same handshape as “WHAT” (i.e., a 5 handshape, palms facing upward),
but it involves a single outward movement, rather than side-to-side shaking of
the hands. In the corresponding video examples, the particle is signed with one
hand (the dominant hand) in (12), but with both hands in (13), as indicated by
the notation ‘(2h).’ Here this is due to the preceding sign being
one-handed (YESTERDAY) or two-handed (CAR). However, there are many variations
that are possible as to whether this particle is signed with the dominant hand,
the non-dominant hand, or both (potentially held over differing
Given the similarity of the articulation, and the fact that this
particle occurs frequently in wh-questions, it is understandable why it might
have been confused with the sign “WHAT” by researchers (including
ourselves) in the past. However, this indefinite particle (unlike the wh-sign
“WHAT”) can occur in non-wh contexts.
3.1. Distribution: Occurrence
in Non-wh Constructions
The particle just described can occur in a wide variety of contexts, the
unifying characteristic of these contexts being some degree of uncertainty.
Along with wh-constructions of the type shown in (12) this particle can occur in
yes-no questions involving some kind of indefiniteness, whereas the wh-sign
“WHAT” would be ungrammatical in yes-no questions.
Hyperlinks: Example 14
Hyperlinks: Example 15
The sign glossed as ‘SOMETHING/ONE’ can be used
to mean either ‘something’ or ‘someone’. The meaning
conveyed by the particle is not reflected in the English translation of these
sentences, as there is no natural way in English to convey this. The semantics
associated with this particle will be discussed in section 4. For now, we will
simply observe that this particle occurs with great frequency in yes-no
questions. Note that it would not be acceptable to replace the particle in (14)
and (15) with “WHAT”.
This particle also occurs frequently in negative constructions, as
illustrated in (16) and (17).
Hyperlinks: Example 16
Hyperlinks: Example 17
Again, replacing the indefinite particle with the sign “WHAT”
in these sentences would result in ungrammaticality.
The indefinite determiner SOMETHING/ONE used in the yes-no question in
(14) is also sufficient to license the indefinite particle even in affirmative
sentences. When the particle occurs with the indefinite determiner/pronoun, it
may appear in any of several positions (and sometimes in more than one
position), as illustrated in (18)–(26). The interpretation differs a bit
depending on where the particle occurs (as we have attempted to show with the
English translations, although, as mentioned earlier, these translations do not
completely reflect the meanings). The particle in (19) is initially signed with
two hands, after which the non-dominant hand retains the position of the
particle throughout the articulation of the sign SOMETHING/ONE on the dominant
hand; the non-dominant hand then participates in the articulation of the
two-handed sign BOAT that follows.
Hyperlinks: Example 18
Hyperlinks: Example 19
Hyperlinks: Example 20
Hyperlinks: Example 21
Hyperlinks: Example 22
Hyperlinks: Example 23
Hyperlinks: Example 24
Hyperlinks: Example 25
Hyperlinks: Example 26
The English translations given above represent the situation
where SOMETHING/ONE refers to an inanimate entity (although, as previously mentioned,
it can also be used with the meaning of ‘someone’). As we see, when
the particle is used sentence-finally, it often indicates uncertainty about
the proposition as a whole, whereas when it modifies a determiner or pronoun
internal to the clause, it indicates uncertainty with respect to the noun phrase
with which it is associated.
This particle also frequently occurs in other constructions that involve
uncertainty, such as sentences that contain non-factive verbs such as
‘guess’ and ‘think’ as in (27) and (28), or adverbials
such as MAYBE, as in (29).
Hyperlinks: Example 27
Hyperlinks: Example 28
Hyperlinks: Example 29
It is also interesting to note in passing the similarity in
articulation between the indefinite particle and the sign MAYBE. MAYBE is
articulated with the same handshape as part:indef: the hands are
extended, with a 5 handshape, palms facing upward; one hand is raised slightly
as the other hand is lowered slightly, and then the hands reverse those motions.
There are other signs with similar meaning and handshape; for example, the sign
meaning ‘approximately’ is articulated with the same handshape, palm
facing forward, and a slight circular movement.
Two additional examples taken from stories by Mike Schlang illustrate
the use of the particle with the verbs HOPE and WISH:
Hyperlinks: Example 30
Hyperlinks: Example 31
In summary, then, this particle can occur in sentences
entailing some degree of uncertainty or tentativeness, whether this is expressed
through verbs or adverbs.
In addition, the particle may occur in situations where the uncertainty
expressed relates to the discourse context, rather than to some specific element
present in the sentence itself. In the literature, the particle occurring in
such contexts has frequently been glossed as ‘WELL’. However, we
believe that most (if not all) occurrences of the discourse particle WELL
actually involve this same particle of indefiniteness.
Hyperlinks: Example 32
Hyperlinks: Example 33
The implicit uncertainty in (33) seems to be in regard to
whether father will really do it. That is, it is significantly less than certain
that father will give the car to
Although many occurrences of the particle traditionally glossed as
‘WELL’ can be understood as involving this particle of uncertainty,
it is not clear whether all instances can be analyzed in this way. There also
seems to be a kind of emphatic use of this discourse-level particle to mark a
transition between topics. In the vast majority of these cases, some type of
uncertainty may indeed be involved, but it is unclear how to succinctly
characterize its distribution with respect to discourse structure.
Many occurrences of this particle are found in excerpts of several
stories distributed by DawnSignPress that we have annotated using SignStream.
These annotations are available on CD-ROM, and we would invite readers to
observe the use of this particle in the specific discourse context of stories,
such as Freda Norman’s “Dead Dog” story. One example is
Hyperlinks: Example 34
Here the uncertainty presumably relates to how this could
There is also another kind of discourse use of this particle, involving
a stressed production of part:indef, separated from the rest of the
sentence by a marked prosodic break, as illustrated in (35) and (36). In
question contexts, this emphatic articulation of the particle conveys a stronger
request for a reply than when the particle is articulated without stress. (The
particle is, in this case, accompanied by certain other nonmanual markings
associated with questions, such as wide eye aperture, raised or lowered
eyebrows—depending on the type of question—and a forward lean of the
body.) This particle (like its unstressed counterpart) can be used in yes-no
questions, unlike the sign “WHAT”.
Hyperlinks: Example 35
Hyperlinks: Example 36
In summary, then, the particle under consideration is articulated in a
way that is similar to, but distinguishable from, the wh-sign glossed as
“WHAT”. The sign “WHAT” involves a side-to-side movement
of the hands, while the indefinite particle involves a single outward movement.
Whereas the sign “WHAT” is restricted to wh-contexts, the indefinite
particle may appear in a variety of other contexts where there is indefiniteness
or uncertainty with respect to an element in the sentence. Given that
wh-questions themselves intrinsically involve an indefinite operator, this is an
ideal environment for the indefinite particle to occur; however, the use of the
indefinite particle is more general. Before returning, in sections 3.4 and 3.5,
to consider other constructions in which this particle occurs, it will be useful
to discuss, in more detail, the realization of this particle in ASL: both its
manual expression and its nonmanual correlates.
In light of the frequency with which this particle occurs, it is rather
striking that it has not been described and analyzed previously. The
phonological reduction that frequently occurs may have been a contributing
factor in its having been overlooked. There is a tendency for this particle to
cliticize phonologically (or to contract) with the sign it follows. In such
cases the handshape and location of the prior sign are maintained, with
part:indef being realized by the addition of an outward movement of one
or both hands (depending on whether it follows a one-handed or two-handed sign).
Often, this kind of assimilation occurs after the wh-signs WHY, FOR-FOR (a
variant meaning ‘why’ or ‘what for’), WHO, HOW,
HOW-MANY, WHERE, WHICH, etc. In the examples below, the symbol ^ represents
such contractions, e.g., WHY^part:indef.
Hyperlinks: Example 37
Hyperlinks: Example 38
Hyperlinks: Example 39
Hyperlinks: Example 40
In (37) through (39), the wh-signs are articulated solely
with the non-dominant hand, which finishes off with an articulation of the
indefinite particle; the articulation of the indefinite particle on the
non-dominant hand begins at the same time as the wh-phrase (i.e., before the
particle is signed on the dominant hand), and continues through the articulation
of the particle on the dominant hand. In (40), however, since HOW is a 2-handed
sign, both hands articulate the wh-sign followed by the particle.
Sometimes, the movement associated with the indefinite particle is very
subtle. For instance, there are cases where it is unclear whether or not the
wh-sign (“WHAT”) is followed by this particle. Frequently the
movement associated with the particle may be subtler on one hand than on the
other. Sometimes, in fact, when the non-dominant hand has not been actively
engaged in signing (but may, for example, be resting on the lap of a seated
signer), there is nonetheless a small yet detectable outward turning of the palm
on the non-dominant hand as the dominant hand produces this particle.
In addition to its ability to cliticize, the particle is also capable of
undergoing perseveration in an utterance. Perseveration (maintenance of a
handshape that will be used again) is a fairly general phenomenon in ASL; for
instance, it occurs frequently with classifiers. Manual perseveration of the
non-dominant hand in the production of the sign “WHAT” was first
noted by Neidle et al. (1994). It was observed that, in constructions involving
two instances of “WHAT” (see note 3), the non-dominant hand may
retain the handshape for “WHAT” between two occurrences of the sign.
Sentence (41) illustrates this phenomenon.
Hyperlinks: Example 41
As in those cases involving “WHAT”, the particle
may also exhibit perseveration by the non-dominant hand throughout part or all
of the utterance, as shown in (42) and (43).
Hyperlinks: Example 42
Hyperlinks: Example 43
There are nonmanual expressions that typically occur with expressions of
uncertainty. As observed by MacLaughlin (1997, p. 119), nouns, verbs, and
adjectives conveying some degree of uncertainty—including the sign
SOMETHING/ONE, which functions as determiner or pronominal—are frequently
associated with a tensed nose, lowered brows, and sometimes also raising of the
shoulders. These same expressions frequently occur with the indefinite particle,
although the eyebrows are sometimes raised (typical of focused constituents).
We have also observed that this particle frequently occurs with a sudden
shift in eye gaze to the left or right, as shown in Figure 2 with respect to
sentences (24) and (21), discussed earlier and repeated here as (44) and
Hyperlinks: Example 44
Hyperlinks: Example 45
Figure 2. Eye gaze occurring with part:indef
This eye behavior remains to be studied carefully, but there
is a definite pattern that has emerged from our data.
Another Construction in Which the Indefinite Particle Occurs
We now return to consider other constructions in which this particle
frequently appears, expressed with the dominant hand, the non-dominant hand, or
both hands. As was pointed out by Lana Cook (personal communication), this
particle often occurs in constructions that have the meaning of ‘according
to’. This is illustrated by the following sentences:
Hyperlinks: Example 46
Hyperlinks: Example 47
Hyperlinks: Example 48
As with the complements of non-factive verbs (recall (27)-(29)),
these are situations in which the truth of the embedded proposition is called
into question. It is interesting that in these examples the particle is signed
solely on the non-dominant hand and may be held, at least until that hand
is needed to participate in the articulation of a 2-handed sign, such as RAIN
Another similar use of the indefinite particle is illustrated
in the following example, taken from a story by Mike Schlang:
Hyperlinks: Example 49
As already noted, this particle occurs with great frequency
in wh-questions, as illustrated here:
Hyperlinks: Example 50
Hyperlinks: Example 51
Hyperlinks: Example 52
Hyperlinks: Example 53
In some cases, the particle may be articulated on one hand while the wh-phrase
is articulated with the other, as occurred in sentence (8), repeated here
Hyperlinks: Example 54
We have identified a previously overlooked particle in ASL
that occurs with great frequency in question constructions and sentences that
involve some kind of indefiniteness. This particle occurs with a precise distribution
and interacts phonologically with other signs, as discussed in section 3.2.
Semantics of the Indefinite Focus Particle
In this section, we will attempt a more rigorous statement
of the contribution of part:indef to the meaning of the utterance.
We note at the outset, however, that our conclusions will necessarily be tentative,
as the full range of contexts in which the part:indef particle can
appear requires further systematic exploration.
Let us consider first the cases in which part:indef
appears with indefinites. Here, the effect of part:indef seems to be
to extend the domain of reference to beyond the typical, resulting in a “widening”
reminiscent of that proposed by Kadmon and Landman (1993) for English any.
That is, whereas (55) simply asserts that a boat sank near Cape Cod, (56)
(repeating (18)) asserts that some (perhaps unusual) kind of boat sank near
Cape Cod, and (57) (repeating (24)) asserts that some boat or perhaps something
only relevantly like a typical boat sank near Cape Cod.
Hyperlinks: Examples 55
Hyperlinks: Example 56
Hyperlinks: Example 57
Adapting Kadmon and Landman’s (1993) condition on any
we could characterize this as follows:
Thus, whatever the extension of the phrase would be without
part:indef (normally fairly restricted contextually), it would be expanded
to include other referents when part:indef is attached. The examples
in (55)–(57) also demonstrate that part:indef has scopal properties;
it can attach to different types of phrases and will widen the interpretation
of whatever phrase it is attached to. In this connection, however, we should
also point out part:indef appears to be allowed at or above its logical
scope position (at least superficially like only in English); thus,
many of the cases discussed below where part:indef appears higher in
the structure also have alternative interpretations in which part:indef
logically associates with an internal constituent.
The explanation of part:indef with indefinites given
above can be extended straightforwardly to the cases of part:indef
with wh-words. We take a wh-word, in a certain sense, to “stand
in for” the possible phrases that could replace it in a well-formed
answer (see, e.g., Hamblin (1973) and most subsequent work on questions).
Usually, the range of values that a wh-word can stand in for is contextually
restricted in much the same way as an indefinite like someone, and
just as with the indefinites discussed above, when used with a wh-word
part:indef also expands the domain of possible referents. This conveys
the feeling that the questioner really has no idea what the answer is, that
the true answer might be outside the set of possible answers the questioner
would consider typical. 
Whereas part:indef appears to be subject to Widening,
it does not appear to be subject to Kadmon and Landman’s (1993) Strengthening
condition on English any. This is clear already from (56): the fact
that some (perhaps unusual) kind of boat sank near Cape Cod does not further
entail that a (usual) boat sank near Cape Cod. However, the Widening effect
on part:indef does make it particularly well-suited for contexts in
which negative polarity items like the English any appear, since it
creates a stronger (more informative) statement. One example illustrating
the semantic Widening effect (here on the main predicate) is given below,
taken from a story by Mike Schlang, “Dorm Prank”. In (59), the
particle extends the characterization of the hall monitor that is being negated:
not only was the hall monitor not friendly, he was nothing like friendly.
Hyperlinks: Example 59
Returning to the cases where the particle is associated semantically
with a sub-sentential constituent, such as WHO or SOMETHING/ONE, it is noteworthy
that the particle is used only when that constituent is in focus. Consider,
for example, the cases where WHO is followed by a phonologically reduced version
of the particle (expressed solely with the dominant hand). This does not naturally
occur when the wh-phrase is in situ. Sentence (60) would be
unnatural (without a great deal of stress on the first sign followed by a
significant pause, marking that in situ phrase as being in focus).
In contrast, the sentences shown in (61) would occur quite
naturally in a context where it was known that somebody saw Joan and the questioner
wished to ask who saw Joan. Again, rightward wh-movement of WHO occurs
only when it is focused.
Hyperlinks: Example 61a
Hyperlinks: Example 61b
Moreover, part:indef is not allowed with just anything
focused in the sentence. For use of part:indef, it must be the indefinite
that is in focus. If the questioner wanted to know whether anyone saw
Joan (that is, in a context like “I know that Bill didn’t
see Joan, but),the questions in (62) are quite natural, but with focus on
Joan (that is, in a context like “I know that nobody saw Bill,
but), neither variant in (62) would be natural.
Hyperlinks: Example 62b
As expected given the analysis so far, we find that in a negative
context, such as (63)(=16)), the particle can play a role similar to English
any. On one reading of (63), the particle serves to emphasize that
mother should not buy any cars, typical or not.
Hyperlinks: Example 63
Another available reading of (63) comes about by Widening
higher, at the VP level, where what is meant is that Mother should not buy
a car or do anything like buying a car. In English, this kind of meaning can
be expressed colloquially with “or something” or “or anything,”
as in the following examples.
Did you go to Europe?
Did you go to Europe or anything?
I didn’t buy a book or anything.
Sentence (64) is simply asking whether or not the proposition
that you went to Europe is true, while (65) asks whether you went to Europe
or did anything like going to Europe. Likewise, (66) is most commonly used
to mean that I didn’t buy a book or do anything that would be considered
in the context to be like buying a book.
In other cases we have seen, part:indef appears to
be attached still higher in the structure, lending a feeling of “uncertainty.”
For example, we might paraphrase (67) (repeating (22)) as ‘That a boat
sank off Cape Cod is likely but not certain’ and (68) repeating (32))
as ‘That John likes cars and books is likely but not certain.’
Hyperlinks: Example 67
Hyperlinks: Example 68
In the normal course of cooperative conversation, a speaker
will say only things that s/he believes to be true, and moreover, in the absence
of any qualification, to be true for sure. If we allow for propositions
to be true to varying degrees of certainty, we can characterize the contribution
of part:indef as serving to allow consideration of degrees of certainty
other than for sure. We can view this as another case of Widening,
now on the degree of certainty with which the speaker regards the proposition
to be true. We can think of “degree of certainty” in a sentence
as being identified with the polarity of the sentence (where in the absence
of a qualifier like part:indef, a negative sentence would be “false
for sure” and an affirmative sentences would be “true for sure”);
the extension here is that under certain circumstances, values between the
two can be explicitly evoked.
The indefinite particle may be used to widen the speaker’s
degree of certainty with respect to a proposition. The sentence in (69) expresses
that it ought to be the case that father will give the car to John, but at
the same time expresses some doubt as to whether, in reality, this will happen.
It is not that the signer isn’t certain about what he or she believes
ought to happen, but rather that the signer isn’t certain that
that father will give the car to John. 
Hyperlinks: Example 69
In these uses (as is clear in some of the other examples,
as well), part:indef functions like a speaker-oriented adverb (akin
to fortunately, certainly, or presumably).
The last function of part:indef we will consider here
is its use in managing conflicts in the discourse. If evidence arises that
participants disagree about the plausibility of a proposition or presupposition,
part:indef can be recruited to assist. Consider (70) (repeating (17)),
uttered in response to a question asking what color John’s car is.
Hyperlinks: Example 70
By asking the question in the first place, the questioner
implicitly indicates that s/he believes the addressee knows the answer. If
this is not the case, one option the addressee has is to respond “I
don’t know,” but (70) goes a step further: “I don’t
know—but it’s not my fault, as I have no way of knowing.”
The way (70) does this is by picking out the prototypical means by which the
answer might be known (i.e., having seen the car) and asserting the falsity
of this proposition and all propositions involving means like it (even perhaps
non-prototypical means, e.g., psychic revelation) by which the answer might
be known. This, too, can be seen as a Widening effect, in this case on the
entire proposition itself (expanding the referent to the proposition and propositions
deemed similar, given the context).
In a related use, consider a situation in which Mary
asks John a question. The presupposition that led Mary to ask John such a
question is her belief that he might know the answer. Suppose that Pete does
not believe that this presupposition is plausible. Pete might say to Mary:
Hyperlinks: Example 71
Here, part:indef is used to communicate the utter falsity
of that presupposition: not only is it false that John knows the answer, anything
(contextually) like John knowing the answer is also false.
It is not entirely clear whether a different kind of analysis
is needed to account for the situations in which the particle seems to be
associated with a presupposition or potential causality that is not explicitly
stated. For example, compare (71) with (72). The meaning contribution of the
particle is essentially the same in both cases; yet, in (72), there is no
explicit negation in the sentence. Each of the examples is a denial of an
implicit presupposition of the addressee. In (72), the addressee’s apparent
belief that John might not know the answer is denied by the signer, with prejudice.
It is clear that part:indef is mostly responsible for the disbelieving
tone of such examples, but a formal solution based on a Widening effect in
the pragmatic context has so far proved elusive.
This is in some ways similar to the example given in (34),
repeated below as (73).
Hyperlinks: Example 73
Here, what is being highlighted is the disconnect between
the indisputable reality and the expected situation. Whether such constructions
should be analyzed as Widening, in a sense related to (but perhaps a bit
different from) the others we have considered, or whether a different account
would be more appropriate for these usages of the particle must be left
as a question for further investigation.
To summarize, it appears that, in most cases, part:indef
semantically associates with some layer of the linguistic structure (a determiner,
a noun phrase, a predicate, sentence polarity, or an entire proposition)
and “widens” the domain of reference, in some direction that
is contextually appropriate. It is worth reiterating that the landscape
of facts is still being explored, making certain details of the analysis
necessarily preliminary; however, the overall pattern seems to fit well
with the approach we have endorsed here.
We have described here a particle that occurs with great
frequency in ASL but which has not been previously analyzed in the literature,
to our knowledge. Its articulation is similar to, but distinguishable from,
the wh-sign glossed as “WHAT.” Although this particle does occur
commonly in wh-questions, it also appears in a variety of other environments
in which wh-phrases are disallowed. Further study of the semantics, distribution,
and use of this particle is warranted. We have argued here that this particle
functions to widen the domain of items referred to by a focused wh-phrase
or indefinite quantifier such as ‘someone.’ This particle may
also be used in a comparable way with respect to other constituents in the
sentence (e.g., NP, VP, CP) and we have suggested that the interpretation
in such cases also involves semantic Widening along a contextually determined
gloss (nearest conventional English equivalent) for
a sign. Proper names in this paper are actually fingerspelled, but,
for ease of presentation, this is not marked explicitly here.
multiword gloss for a single sign
a slash is used when a single ASL sign has more than
one English translation
a sign in ASL that may be translated as either ‘something’
indicates that a sign is fingerspelled (H-A-L-L)
indicates a gloss for a fingerspelled loan sign.
^ indicates contraction
a wh-sign produced with both hands extended, palms
facing up, moving slightly from side to side
an indefinite focus particle articulated with a single
outward movement of one or both palms (facing upward)
pointing sign (used for pronominal reference); marked
for first person
pointing sign (used for pronominal reference); non-first
a possessive pronoun, articulated with an open palm
pointing toward the possessor
extended line indicates the domain over which the
non-manual marker occurs
wh-question marker (furrowed brows, squinted eyes,
sometimes accompanied by a slight side-to-side head shake)
yes-no question marker (includes raised eyebrows,
forward head tilt)
negative marking (consisting of a side-to-side head
shake and furrowed brows)
trace coindexed with some moved element
dominant and non-dominant hands
Normally, the glosses in this article do not reflect
whether the sign is 1-handed or 2-handed. The only exception to this
is the part:indef, which is glossed as (2h)part:indef if and only
if it is signed with 2 hands.
In some cases, where the behavior of the non-dominant
hand is, to some degree, independent of what is signed on the dominant
hand, separate lines are used in the gloss to represent the unusual
behavior of the non-dominant hand. The labels of [d] for ‘dominant
hand’ and [nd] for ‘non-dominant hand’ are used,
but in fact, the first line still may contain some predictable information
about the use of the non-dominant hand.
In this case, JOHN is a fingerspelled name sign, signed
only with the dominant hand; LOVE is a two-handed sign. The divergence
of dominant and non-dominant hand behavior is only as indicated by
information on both lines. In this case, the particle is articulated
first with the non-dominant hand: at the same time that the dominant
hand signs WHO. Both hands complete the articulation of the particle
at the same time.
Baker, Charlotte, and Dennis Cokely. 1980. American
Sign Language: A Teacher’s Resource Text on Grammar and Culture. Silver
Spring, MD: T.J. Publishers.
Baker, Charlotte, and Carol A. Padden. 1978. Focusing
on the Nonmanual Components of American Sign Language. In Understanding
Language through Sign Language Research, edited by Patricia Siple, 27-57.
New York: Academic Press.
Baker-Shenk, Charlotte. 1983. A Micro-Analysis of
the Nonmanual Components of Questions in American Sign Language. Doctoral
Dissertation, University of California, Berkeley.
den Dikken, Marcel, and Anastasia Giannakidou. 2002.
From Hell to Polarity: “Aggressively Non-D-Linked” Wh-Phrases
as Polarity Items. Linguistic Inquiry 33, no. 1: 21-61.
Emmorey, Karen. 1999. Do Signers Gesture? In Gesture,
Speech, and Sign, edited by Lynn Messing and Ruth Campbell, 133-59. Oxford:
Oxford University Press.
Gärdenfors, Peter. 1988. Knowledge in Flux:
Modeling the Dynamics of Epistemic States. Cambridge, MA: MIT Press.
Hamblin, C.L. 1973. Questions in Montague English.
Foundations of Language 10: 41-53.
Kadmon, Nirit, and Fred Landman. 1993. ANY. Linguistics
and Philosophy 16: 242-422.
Lillo-Martin, Diane, and Susan Fischer. 1992. Overt
and Covert Wh-Questions in American Sign Language. Salamanca, Spain: Fifth
International Symposium on Sign Language Research.
MacLaughlin, Dawn. 1997. The Structure of Determiner
Phrases: Evidence from American Sign Language. Doctoral Dissertation, Boston
University, Boston, MA. (available from http://www.bu.edu/asllrp/)
Neidle, Carol. in press. Language across Modalities:
ASL Focus and Question Constructions. Linguistic Variation Yearbook 2.
———. 2002. SignStream™Annotation:
Conventions Used for the American Sign Language Linguistic Research Project.
American Sign Language Linguistic Research Project Report No. 11, Boston
University, Boston, MA. (available from http://www.bu.edu/asllrp/)
———. 2001. SignStream™ A
Database Tool for Research on Visual-Gestural Language. Journal of Sign
Language and Linguistics 4, no. 1/2:203-214.
———, ed. 2000. ASLLRP SignStream
Databases CD-ROM, Vol. 1. Boston, MA: American Sign Language Linguistic
Research Project, Boston University, Boston, MA.
Neidle, Carol, Judy Kegl, and Benjamin Bahan. 1994.
The Architecture of Functional Categories in American Sign Language. Talk
presented at Harvard University, Cambridge, MA, May 1994
Neidle, Carol, Judy Kegl, Benjamin Bahan, Debra Aarons,
and Dawn MacLaughlin. 1997. Rightward Wh-Movement in American Sign Language.
In Rightward Movement, edited by D. Beerman (sic), D. LeBlanc and H. Van
Riemsdijk, 247-78. Philadelphia: John Benjamins.
Neidle, Carol, Judy Kegl, Dawn MacLaughlin, Benjamin
Bahan, and Robert G. Lee. 2000. The Syntax of American Sign Language: Functional
Categories and Hierarchical Structure. Cambridge, MA: MIT Press.
Neidle, Carol, Stan Sclaroff, and Vassilis Athitsos.
2001. SignStream™ A Tool for Linguistic and Computer Vision Research
on Visual-Gestural Language Data. Behavior Research Methods, Instruments,
and Computers 33, no. 3: 311-20.
Nilsen, Øystein. to appear. Domains for Adverbs.
Pesetsky, David. 1987. Wh-in-Situ: Movement and Unselective
Binding. In The Representation of (in)Definiteness, edited by Eric Reuland
and Alice G.B. ter Meulen, 98-129. Cambridge, MA: MIT Press.
Petronio, Karen, and Diane Lillo-Martin. 1997. Wh-Movement
and the Position of Spec-CP: Evidence from American Sign Language. Language
73, no. 1: 18-57.
Rizzi, Luigi. in press. Locality and Left Periphery.
In Structures and Beyond: The Cartography of Syntactic Structures, volume
3, edited by A. Belletti. Oxford: Oxford University Press.
———. 1990. Relativized Minimality.
Cambridge, MA: MIT Press.
Romero, Maribel, and Chung-hye Han. 2001. On Certain
Epistemic Implicatures in Yes/No Questions. In Proceedings of the 13th Amsterdam
Colloquium. Amsterdam: ILLC/Department of Philosophy, University of Amsterdam.
van Rooy, Robert. to appear. Attitudes and Context
Change. Manuscript to be published by ILLC/Department of Philosophy, University
is based upon work supported by the National Science Foundation under grants
SBR-9410562, BCS-9729010, IRI-9528985, IIS-9912573, IIS-0329009 and EIA-9809340.
We are especially indebted to Ben Bahan, Mike Schlang, and Lana Cook for
sharing their intuitions and ideas about this particle. This work has also
benefited from discussions with Robert G. Lee, Sarah Fish, Carla DaSilva,
Dawn MacLaughlin, Norma Bowers Tourangeau, and Jean Berko Gleason. We are
grateful to Stan Sclaroff, Vassilis Athitsos, Murat Erdem, and Sarah Fish
for their assistance in production of the video files illustrating the example
sentences in this article. Authors’ names for this article are listed
 See, e.g.,
Petronio and Lillo-Martin (1997) for different claims about ASL questions
from those presented by Neidle et al.(1997, 2000, e.g.) and summarized in
 The distribution
of nonmanual markings has been carefully analyzed using SignStream, a program
designed to facilitate the linguistic analysis of visual language data by
providing tools for on-screen display and analysis of linguistic information
alongside of the actual video. SignStream is distributed on a non-profit
basis to students, educators, and researchers, and the coded data are also
publicly accessible. See <http://www.bu.edu/asllrp/SignStream/>.
For further information about the Center for Sign Language and Gesture Resources,
through which we have been collecting and distributing high quality video
data (four synchronized views) plus linguistic annotations thereof (Neidle,
ed. 2000), see Neidle, Sclaroff and Athitsos (2001) and <http://www.bu.edu/asllrp/cslgr/>.
 As discussed
in Neidle et al. (2000:197 fn.10), multiple wh-questions are generally disallowed,
except when the wh-phrases are strongly D-linked (Pesetsky 1987). However,
there are constructions that involve more than one occurrence of a wh-phrase
associated with a single questioned argument: an initial wh-phrase (which
we have analyzed as a kind of topic) followed by a clause that contains
a wh-phrase either in situ or at the right periphery.
 Thus, for
example, the sentences in (5)-(8) have a presupposition that there was someone
who was seen or who arrived, analogous to English sentences with intonational
focal stress on ‘who’ in ‘Who arrived?’ or
‘Who did John see yesterday?’. See Neidle (in press)
for an analysis in terms of leftward focus-movement followed by rightward
wh-movement—subject to Relativized Minimality (Rizzi, e.g., 1990,
in press)—in the derivation of sentences where the wh-phrase surfaces
at the right periphery.
 See, e.g.,
Baker and Cokely (1980), Baker and Padden (1978), or Baker-Shenk (1983).
 This distinction
was first observed by Lillo-Martin and Fischer (1992).
 There are
two other ASL signs (which occur less frequently than “WHAT”)
that can be used with the meaning of the English word ‘what’:
#WHAT, a loan sign in which the English word is rapidly fingerspelled w-h-a-t,
and another sign, often glossed as ‘WHAT’ (with single rather
than double quotation marks), which is produced by brushing the index finger
of one hand across the fingers of the other, open 5 hand. There are some
differences in the distribution of these signs that have yet to be fully
described (see Neidle et al., 1997:262 and fn. 24).
 In addition
to the particle shown in these examples at the end of the sentence, it is
also possible to have a second occurrence of the particle immediately following
the verb THINK or SEEM (sometimes with a pause after the particle).
 As pointed
out by Lana Cook (personal communication), this would include the situation
where father died and, therefore, where it has become impossible for him
to give the car to John. What is crucial is that there is at least the
possibility that father will not give the car to John.
 In Emmorey
(1999), this particle is considered to be a gesture and glossed as ‘/well-what/.’
Although it is not entirely obvious how the distinction between sign and
gesture should be defined, even by the criteria provided by Emmorey, part:indef
does seem to pattern with linguistic signs (rather than gestures) by
virtue of systematicity of form, distribution, and meaning.
in this respect, part:indef is similar to “...he hell”
in English wh-questions (as discussed by Pesetsky (1987), den Dikken
and Giannakidou (2002)), it does not appear to be as “aggressive”
in its domain-widening.
 This type
of “Bayesian belief model” has been explored by several authors;
see, e.g., Romero and Han (2001), Nilsen (to appear), van Rooy (to appear),
and particularly Gärdenfors (1988).
 This is
perhaps similar to the English “Father certainly must give the car
to John” (noting that, perhaps counter-intuitively, one of properties
of the speaker-oriented adverb certainly is that it introduces a
small degree of doubt; compare ‘Audrey died in the explosion’
with ‘Audrey certainly died in the explosion’).