Language Sciences 50 (2015) 12–29
Contents lists available at ScienceDirect
Language Sciences
journal homepage: www.elsevier.com/locate/langsci
Expressive meaning in an AAE attributive constructionq
Patricia Irwin
University of Pennsylvania, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 1 September 2014
Received in revised form 6 January 2015
Accepted 6 January 2015
Available online
This paper presents an overview of the syntax, semantics, and prosody of the discourse
-ass construction in African American English, as in get all that ugly-ass junk out of here.
This construction involves attributive modification in which a noun or adjective (called the
ASSOCIATE) forms a constituent with the word ass and modifies a head noun. The paper
describes the syntactic distribution of both the ASSOCIATE and the word ass. Arguments are
presented that support an analysis in which -ass is not a nominal but a functional head that
categorizes its sister as adjectival, similar to -ish and -y in mainstream English. Semantically, it is argued that discourse -ass is an expressive in the sense of Potts (2007b): it is
“semantically bleached” (Spears, 1998), and its semantic contribution is not truth conditional. The paper shows how discourse -ass has the properties associated with expressives
as articulated by Potts (2007b) and as first observed about the construction in Spears
(1998).
Ó 2015 Elsevier Ltd. All rights reserved.
Keywords:
African American English
Expressive meaning
Grammaticalization
Morpho-syntactic variation
Swear words
1. Introduction
Colloquial African American English has a construction that has been noted by Spears (1998), Collins et al. (2008), and
others, and which is exemplified and underlined in (1):
(1)
Get all that ugly-ass junk out of here. (Spears, 2001: 231; underline added)
I will refer to this construction as ‘discourse -ass’, following Spears (1998). This term emphasizes one of the main points of this
article, that this construction is expressive, in the sense of Potts (2007b). Spears (1998) was the first to argue that the meaning
of discourse -ass is its expressive contribution to a context, and the first goal of this paper is therefore to expand on these
observations in the context of recent work on expressive meaning, primarily as articulated by Potts (2007b). The second goal
of the paper is to lay out the syntactic distribution of the construction. This aspect of the paper is important, because if the -ass
in discourse -ass makes no meaningful contribution to the construction on the truth-conditional semantic level, then we must
look to the syntax to determine the constraints on its distribution. Although the phrase ugly-ass may appear at first to have the
structure of a compound, this paper will present arguments that support an analysis in which the -ass of discourse -ass
patterns more like bound affixes such as -ish and -y in English, or -(c)ito/-(c)ita in Spanish, than as part of a compound.
In the prototypical case, a discourse -ass construction consists minimally of two parts: an adjective or noun (e.g., ugly, fool),
and the word ass. Together, these two parts modify a noun, as shown in (1), where ugly ass modifies junk. The structure of
discourse -ass is shown schematically in (2), which gives the terms that I will use in analyzing the construction.
q Thanks to Chris Collins, Richard Kayne, John Singler, and Arthur Spears for comments on earlier versions of this paper. Thanks to the following people
for support and feedback on this version: Gene Buckley, Adam Croom, Dave Embick, Sabriya Fisher, Neil Myler, Florian Schwarz, and Jim Wood. This
research was assisted by a New Faculty Fellows award from the American Council of Learned Societies, funded by The Andrew W. Mellon Foundation.
E-mail address: [email protected].
http://dx.doi.org/10.1016/j.langsci.2015.01.002
0388-0001/Ó 2015 Elsevier Ltd. All rights reserved.
P. Irwin / Language Sciences 50 (2015) 12–29
(2)
13
ugly
-ass
junk
a-word HEAD-N
MODIFIER
]
ASSOCIATE
[
I refer to -ass in this construction as the a-word, since in some cases it can be something other than -ass, and to minimize the
distraction of repeatedly using a swear word. The pre a-word material (e.g., ugly) will be referred to as the ASSOCIATE, and we
will see that although the associate is typically an adjective or noun, it is not confined to these syntactic categories. I will refer
to the N in the construction as the “head noun” (HEAD N). Following Spears (1998), I will often hyphenate the associate and the
a-word in order to indicate that they form a prosodic constituent; adopting this typographic convention can also help
disambiguate uses that involve the word ass but do not have the structure or meaning of discourse -ass.
As shown in (2), I will use the descriptive term MODIFIER to refer to the constituent that consists of the associate and the aword. In addition, I will sometimes refer to the modifier as the “associate þ a-word” constituent, particularly when discussing
the modifier as two lexical items.1
On its own, the word ass means ‘buttocks’, and it is considered a mild to moderately strong swear word in African
American English (AAE) and most other varieties of English. In the discourse -ass construction, the literal meaning of the word
is absent. This is seen in (1) in the fact that it is used to describe “junk,” which does not denote a being with anatomical
properties. The discourse -ass construction does, however, retain the force of the swear word, and usages like (1) may be
considered impolite or even “obscene” in many speech situations (Spears, 1998).
Because this paper will focus somewhat narrowly on the syntactic and semantic distribution of discourse -ass, it will not
address how the construction fits into the broader context of African American communicative practices and “black ways of
speaking” (Spears, 2007: 226). For insightful discussions on these topics, particularly with respect to discourse -ass, see Spears
(1998, 2001, 2007), and Smitherman (1994, 2000).
1.1. A note on the data
The construction discussed in this paper is part of AAE. Two useful sources for readers unfamiliar with the language are
Green (2002), for an overview of the grammar and phonology of AAE, and Smitherman (1994, 2000) and Spears (1998) for
definitions and information on particular lexical items. The grammaticality judgments in this paper come from informants
who identify as African-American and native speakers of AAE. The two main informants were a 28-year-old male and a 44year-old male. Both informants were life-long residents of the New York City area. The 28-year-old informant had spent 5
years outside the NYC area, in Raleigh, North Carolina. In order to minimize the influence of Caribbean dialects or other
languages and dialects that are spoken primarily outside the U.S., only informants whose parents were born and raised in the
U.S. were used.2 The majority of the judgments in this paper come from these two informants. A third, 50-year-old female
informant participated, but she dropped out part-way through informant work because she did not like giving judgments on
what she judged to be incorrect English, and she was not comfortable with the swear words in the data.
The data come from numerous sources: contrived sentences; examples from previous scholarly work such as Spears
(1998) and Collins et al. (2008) (linguistic examples from this source will be abbreviated ‘CMP’); examples from the
Internet; from the social networking tool Twitter; from popular media such as films and rap songs; and examples offered by
informants. Because it is difficult to determine whether any given example from the Internet is from a speaker of AAE, all the
examples in this paper, regardless of source, were judged by native speaker informants. In other words, although Web pages
and Twitter feeds are cited for some of the data in this paper, the grammaticality judgments on them come from native
speaker informants. Judgments on data from mainstream English are my own.
Much of the data comes from the stand-up comedy film, The Original Kings of Comedy (Lee, 2000). This film contains the
sets of several different comedians. For all examples from this source, the name of the performer who uttered the example
(e.g., Bernie Mac) is given in addition to the conventional (producer, year) citation.
1.2. Distinguishing discourse -ass from similar constructions
Before proceeding to the syntactic and semantic distribution of discourse -ass, we must first distinguish discourse -ass
from superficially similar constructions. The uses discussed in this section all involve the word ass but are distinct from
discourse -ass, and although some of them are likely related to discourse -ass, the connections between them and discourse
-ass must be set aside for future research.
The most closely-related construction is exemplified in (3), which I will refer to as the [þhuman] construction. Using the
terms outlined above, we can say that the [þhuman] construction involves an associate (again, usually an adjective or a noun),
and the word ass. This construction appears to lack a head-N, but here the associate þ ass constituent functions as an
argument.
1
Spears (1998) uses the term “-ass word” (abbreviated to AW) to refer to what I call the MODIFIER or “associate þ a-word” constituent. In Spears (2001),
he refers to AWs as “ass compounds.”
2
This may be a challenge particular to work on AAE in the northeastern United States, though see Spears (2001) for some discussion of commonalities
between African American and Caribbean language practices.
14
P. Irwin / Language Sciences 50 (2015) 12–29
(3)
a. Gwen’s husband is a stupid ass.
b. Some stupid ass is parked in my spot.
This may be called the [þhuman] construction because it can only be used to denote a person, as shown in (4).
(4)
The [þhuman] discourse -ass construction
a. *Jamie’s cat is a lazy ass.
b. *Jamie’s coat is a raggedy ass.
The [þhuman] construction admits a much smaller range of associates than discourse -ass or the “ass camouflage construction” (discussed below).
Another important and likely related construction has come to be called the “ass camouflage construction” (5). It has been
discussed most prominently by Spears (1998), Beavers and Koontz-Garboden (2006), and Collins et al. (2008).
(5)
the ass “camouflage” construction (underlining added)
a. His ass is gonna get fried. (Spears, 1998: 235)
b. Get your triflin’ ass out of here. (Spears, 1998: 235)
c. Stacy ass gon get fired. (CMP, 35)
In this construction, the a-word functions like a reflexive, such as self (Spears, 1998: 235). This construction involves a
possessor (e.g., his, Stacy) and an a-word, which may or may not be modified (e.g., triflin’), and which Collins et al. (2008)
refer to as the “mask.” The whole constituent can serve as an argument. Semantically, the construction denotes the
possessor; in other words, Stacy ass denotes Stacy. As Collins et al. (2008, 30) put it, “the truth-functionally relevant
meaning of a camouflage structure is simply that of its possessor.” One reason why the ass camouflage construction is of
linguistic interest is because the possessor acts syntactically as though it is outside the possessive structure (with respect
to binding, for example), but the a-word determines important syntactic properties like finite verb agreement (Collins
et al., 2008: 30). For a detailed analysis of the camouflage construction, see Collins et al. (2008) and Collins and Postal
(2012).
A construction in AAE that appears to be a variant of the camouflage construction is discussed briefly in Spears (2007),
where a word referring to a different body part (e.g., the eyes) is used (6).
(6)
Get your frog eyes out of heredwhat are you looking at?
Spears writes that “the body part word used in such expressions depends on the discourse context and must be relevant in
some way to the situation” (Spears, 2007: 230).
As all authors on the topic have noted, many varieties of English accept constructions like the ass camouflage construction,
as in examples like (7); Collins et al. (2008) focus the variant in AAE which has a wider and more complex syntactic distribution than uses like those in (7).
(7)
Mainstream ass camouflage
a. The Angels really kicked the Orioles’ ass yesterday. (CMP, 29–30)
b. Get your ass in gear!
Another use of ass that we must set aside is perhaps more overtly expressive than the ones discussed above. This is the
fixed phrase my ass, used across varieties of English to express disbelief about a previous statement that is under discussion
(8).
(8)
Mainstream English: my ass as an expression of disbelief
a. Monsters University is Rated G? My ass it is.3
b. Lmao good answer my ass!4
In the English construction, the previous statement can be made by the speaker herself, or it can be made by an interlocutor.
Interestingly, the phrase my ass seems not to be able to stand alone as an expression of disbelief, making it unlike similar
expressions like no way!. In my dialect, it must attach to a larger constituent, for example a CP with VP ellipsis (8a), or to a
constituent that raises the matter that disbelief is being expressed aboutde.g., good answer in (8b). Despite the interesting
syntactic and pragmatic properties of the expressive my ass, we must leave it for future research.5
There are numerous other constructions across American English dialects using ass, as in the A/OFF construction, mentioned
in Collins et al. (2008), as exemplified in (9).
3
http://www.chicagonow.com/baby-sideburns/2013/07/monsters-university-rated-g/.
http://www.datehookup.com/Blog-611498.htm.
The phrase mon cul ‘my ass’ also occurs as an expression of disbelief in colloquial French, as in L’amitié? Mon cul ouais!! ‘Friendship? Yeah my ass!
(http://vlout.skyrock.com/2096135553-L-amitie-Mon-cul-ouais.html), although this phrase is probably a shortened version of et mon cul c’est du poulet! lit:
‘and my ass is chicken!’. (Thank you to Vincent Chanethom for discussion of the French data.)
4
5
P. Irwin / Language Sciences 50 (2015) 12–29
(9)
15
That guy been working his ass off. (Collins et al., 2008: 31)
Finally, it should be noted that ass as a swear word in British and American English dialects occurs in numerous epithets, from
the mild terms ass, jackass, meaning a foolish or unpleasant person, and smart-ass, to the much stronger asshole, ass-wipe. The
word also occurs in various phrases of rejection or disparagement, as in what might be called the up x’s ass construction (10).
(10)
You can take your rebate certificates and shove them right up your ass!6
Uses like the epithet asshole are expressive in the technical sense: we expect them to pattern with words like bastard in the
properties that Potts (2007b) identifies. But I would argue that unlike discourse -ass, these uses retain some of the literal
meaning of the word. In many of the epithet cases, the anatomical meaning is relevant in its relation to excrement, as seen
quite directly in uses like (11), where the swear word shit can be substituted for ass with no detectable change in meaning.
(11)
Remember when Lebron got 5 fouls and played like ass in game 5 against Indiana?7
The connection to anatomical ass is also seen in the oddity of sentences like (12a), in contrast to a similar sentence with
discourse -ass (12b).
(12)
a. ???Your cat is an asshole.
b. Your stupid ass cat didn’t make it to the litterbox.
Examples like these show that the ass terms above have a [þhuman] feature in addition to having an anatomical feature (since
(12a) is bad, even though cats have rear ends and can be unpleasant).
In discourse -ass, the word has lost all connection to the literal meaning of buttocks. For this reason, work on the construction (as well as the ass camouflage construction), describe the a-word as having been “semantically bleached” (Spears,
1998: 236). This does not mean that the a-word in discourse -ass does not have the force of a swear word, however: it most
definitely does. Discourse -ass derives its expressive force from the fact that it contains a swear word for a body part, though
none of the literal meaning of the swear word is retained. Semantically, this makes discourse -ass a nearly pure example of
expressive language. I suggest below that the semantic bleaching of ass is related to its grammaticalization into a morpheme
with the properties of a functional affix.
1.3. Prosody and discourse -ass
One important property of discourse -ass is that the associate and the a-word form a prosodic unit, and within this unit
(the modifier), the a-word is prosidically subordinated to the associate. This fact was first observed by Spears (1998: 236).
Consider the example in (13).
(13)
We some punk ass parents, see. (Lee, 2000, Bernie Mac)
Cf. We are lenient parents
The pronunciation of the modifier punk ass is shown informally in (14), with primary stress indicated by an acute accent, and
secondary stress indicated by a grave accent.
(14)
púnk àss
The complete phrase has stress on the head N, as shown in (15a). Any other stress pattern, e.g., (15b), is simply infelicitous.8
(15)
a. púnk àss párents
b. ??? púnk áss pàrents
To the extent that (15b) has any meaning, its pronunciation signals a syntactic structure that is different from that in discourse
-ass, one in which the a-word attaches to the head N rather than the associate.
The nominal epithet ass that occurs in nearly all varieties of English, referring to a foolish or unpleasant person and
discussed in x1.2, function, as non-head of a N–N compound. In this morpho-syntactic context, the word ass is stressed. The
politician Sarah Palin provided an example of this construction with the phrase ass-clown on Twitter (16).9
(16)
6
The rest of America is out there working our asses off while these DC assclowns throw themselves a
nerdprom10
http://www.reddit.com/r/nosleep/comments/2a1qk1/you_can_take_your_rebate_certificates_and_shove/.
http://www.qaster.com/q/474763855285874688/.
8
There can be ambiguity between the [þhuman] construction and discourse -ass in sentences like Look at his fat ass. See Spears (1998: 235) and Spears
(2007: 231) for discussion on disambiguating examples like these.
9
Although N–N compounding in English is extremely productive, the productivity of compounds with ass as the non-head appears to be limited.
10
Hashtags removed. The complete tweet is as follows: “That #WHCD was pathetic. The rest of America is out there working our asses off while these DC
assclowns throw themselves a #nerdprom” (https://twitter.com/SarahPalinUSA/status/328346466567479296; posted 8:14 PM – 27 Apr 2013).
7
16
P. Irwin / Language Sciences 50 (2015) 12–29
Cases like ass-clown follow the rules of English compound pronunciation, in that the word ass is stressed, as shown in
(17).
(17)
áss clòwn
Turning back to discourse -ass, one conclusion that might be drawn from the stress pattern of the modifier and the fact that
ass can in other contexts function as a nominal is that the modifier has the structure of a compound word. On this analysis, the
discourse -ass construction in (18a) would have the structure of the compound in (18b).
(18)
Schematic compound analysis
a. [ punk ass ] parents
b. [ bitter sweet ] memory
A compounding analysis is tempting, since compounds in English always have stress on the left-hand element, as shown with
a N–N and a synthetic compound in (19).
(19)
a. gárbage màn
b. bée-keèper
There are several facts that cast doubt on a compounding analysis, however. One fact concerns the stress pattern of the
associate þ -ass constituent: this stress pattern is not only that of English “compound stress”; it is also the stress pattern of
words like those in (20) with derivational affixes, such as -ish and -y.
(20)
a. gréen-ìsh
b. stámmer-ỳ11
I present more detailed arguments against a compounding analysis in x2.5.1, arguments that do not involve prosody. The role
of prosody in discourse -ass is important, howeverdnot just in helping to diagnose the constituent structure of the construction. In many cases, the modifier consists of a metrical foot that is trochaic, and this fact plays a role in how discourse -ass
is used by speakers in African American communicative practices more broadly.
2. Syntactic distribution
The following subsections describe the syntactic properties of discourse -ass, focusing first on the properties of the
associate in the modifier. We will see that the associate is not limited to a single grammatical category, although Spears (1998)
suggests that it is most productive with nominal associates. The associate cannot be inflected. When the associate is an
adjective, only attributive adjectives are acceptable. The modifier (associate þ a-word) cannot occur predicatively, even if the
associate on its own can be used as a predicate adjective or nominal.
We turn next to the status of the a-word. Despite the fact that in its literal usage, the word ass is a nominal referring to
a body part, the a-word in discourse -ass patterns like a functional morpheme in its syntactic and semantic distribution.
The a-word does not function like a nominal in discourse -ass: it cannot be inflected (as noted above), and it does not refer
to an entity in the discourse. Although a complete morpho-syntactic analysis of discourse -ass is beyond the scope of this
paper, I propose that the facts best support an analysis along the lines of the tree in (21), illustrated with the modifier foolass.
(21)
fool-ass
The analysis is shown here with a syntactic approach to word-formation, and -ass is analyzed as an affix
similar to -ish in categorizing its sister as adjectival. Before merging with -ass, the sister may be nominal, as in (21),
adjectival, or a syntactically complex category (which may well be a compound; see Lieber, 1992; Harley, 2009). The
resulting aP structuredthe modifier, in the terms used heredcan modify a nominal in the extended projection of a
DP just as an adjective like foolish would. This is illustrated in (22) with one possible implementation of adjectival
modification.
11
See Adams (2004: 67ff).
P. Irwin / Language Sciences 50 (2015) 12–29
(22)
17
my fool-ass cousin
I refer to this analysis as the little-a analysis, and at the end of this section, it is contrasted with a compounding analysis.
2.1. Attributivity and the properties of the associate
Discourse -ass is productive with a variety of categories in the associate position. As Spears (1998) shows, the associate
of the a-word is not limited to one category: it can be an adjective, a noun, a participle, or even a phrase. The first sections
below will focus on the properties of adjectival associates before turning to nominal, participial, and phrasal associates.
Modifying words are traditionally divided into two types: attributive and predicative (Huddleston and Pullum, 2002).
Predicative modification requires the use of the copula (23a); attributive modification does not (23b).
(23)
Predicative and attributive modification (examples based on Huddleston and Pullum, 2002: 57)
a. That sweater is raggedy.
predicative
b. That raggedy sweater should be thrown out.
attributive
Most adjectives in English can be used either attributively or predicatively, as we see with raggedy in (23). In addition,
attributive adjectives in English usually occur before the head noun, as we see in (23b) (Huddleston and Pullum, 2002, 439).
One striking syntactic fact about modifier of discourse -ass is that it can only be used for attributive modification. In other
words, the associate þ a-word constituent can only occur in a pre-head-N position. Predicative modification by the
associate þ a-word is sharply ungrammatical (24).
(24)
a. That raggedy ass sweater should be thrown out.
b. *That sweater is raggedy ass.
In English, some adjectives that in their attributive use occur in a pre-head position can also occur in a post-head position if
the adjective has a complement. One example of this is proud. When proud does not have a complement, it must precede the
head N; but when it has a complement, it can occur after the head N (25) (see Huddleston and Pullum, 2002: 551–552 for
discussion).
(25)
a. A proud man stood up
b. *A man proud stood up
c. A man proud of his son stood up
In a discourse -ass context, an adjective that normally can take a complement, like proud, can no longer take a complement. Consider proud in (26).
(26)
I’m a proud ass Knick fan, proud of our history.12
It does not matter whether the complement material occurs before or after the a-word. Moving the modifier to a post-head
position does not improve grammaticality.
(27)
12
a. *a father proud ass of his son
b. * a father proud of his son ass13
http://www.knicksonline.com/forums/showthread.php?p¼52070.
This phrase is not grammatical as discourse -ass, but it could be grammatical if it is the construction discussed by Collins et al. (2008)de.g., John is proud
of his son ass z ‘John is proud of his son’.
13
18
P. Irwin / Language Sciences 50 (2015) 12–29
One interesting exception is when the discourse -ass NP does not serve as an argument, as in (28), involving what Collins et al.
(2008: 52) call the “resumptive with” construction.14
(28)
I saw her with her all proud of her son ass at the graduation.
As it stands, however, (28) may be ambiguous between discourse -ass and CMP’s ass camouflage.15 Further informant research
will tell whether (28) has the structure in (29) or in (30), or whether it can have either structure.
(29)
with [DP her [nP [aP all proud of her son] [n ass] ] ]
(30)
with [DP her ] [aP [all proud of her son] [a -ass] ]
camouflage
discourse -ass
The labels in (29) and (30) should be taken with a grain of salt, since as Collins et al. (2008) remark, there is no existing
analysis of the AAE resumptive with construction. One bit of data relevant to whether (30) really is discourse -ass is the fact
that saying (30) would be akin to saying something like (31), where [with her all happy] is an adjectival phrase postnominally modifying a direct object pronoun.
(31)
a. I saw her with her all happy.
b. I saw her with her happy.
Further informant work will determine whether (31) is acceptable in AAE, but it is worth noting that a post-nominal modifier
of a pronoun is acceptable in mainstream English in a context like (32).
(32)
I saw her all happy.
Whatever the status of (28), the fact remains that the associate cannot generally have a complement, and this is related to the
uninflectability of the associate.
2.1.1. No inflection of the associate
One striking feature of the associate in the discourse -ass construction is the fact that it cannot be inflected. For example,
adjectives like cute that can normally be inflected into comparative and superlative forms (cuter, cutest) cannot be so inflected
when they are discourse -ass associates.
(33)
No
a.
b.
c.
inflection on the associate
This my cute ass friend.
*This my cuter ass friend.
*This my cutest ass friend.
The associate þ a-word constituent allows for periphrastic comparative and superlative forms, but these morphemes can only
modify the whole associate þ a-word constituent (34a). Comparative/superlative modification of just the associate is ungrammatical, as shown in (34b).
(34)
my most raggedy ass sweater
a. my [ most [raggedy ass] ] sweater
b. *my [ [most raggedy] ass ] sweater
The uninflectability of the associate is one way in which the discourse -ass modifier appears to pattern like a compound.
But this diagnostic does not distinguish a compounding analysis from a little-a analysis, since derivational morphology in
English cannot attach to comparative/superlative morphology, as shown in (35).16
(35)
a. cute
b. cuteness
c. *cuterness
Put differently, the uninflectability of the associate is predicted by both a compounding analysis and an analysis in which -ass
is an inflectional, category-determining morpheme.
2.1.2. Semantic classes of adjectival associates
Before turning to non-adjectival associates, we will first explore the classes of adjectives that can occur as associates. We
will see that discourse -ass associates allow for a wide range of types of adjectives.
14
15
16
Thank you to Arthur Spears (p.c.) for this observation and the data in (28).
Thank you to an anonymous reviewer for pointing this out.
Thanks to an anonymous reviewer for bringing this to my attention.
P. Irwin / Language Sciences 50 (2015) 12–29
19
We have seen that discourse -ass allows adjectives that can normally take complements and that allow comparative and
superlative forms, but that these adjectives lose these properties when they become discourse -ass associates. Discourse -ass
allows for both gradable (36) and non-gradable adjectival associates (37).
(36)
Gradable associates
a. Scary ass sight ain’t it? (Lee, 2000, Cedric the Entertainer)
b. What you laughing at with them thick ass glasses on? (Lee, 2000, D.L. Hughley)
(37)
Non-gradable associates
a. Parents please tell y’all sick ass children to cover they mouth when they cough!!!17
b. but im not for no red ass car to make my insurance high18
Discourse -ass is also productive with what are called non-subsective adjectives, such as former and fake (see Kamp and
Partee, 1995). Non-subsective adjectives are those that are semantically neither intersective nor subsective; in other
words, we cannot assume that the denotation of the non-subsective adjective–noun combination is a subset of the denotation
of the noun (38). The adjective potential is subsective: if someone is a potential NFL player, we cannot conclude that the
person is an NFL player (though they might be).
(38)
Non-subsective
EAdj NF=ENF
A special type of non-subsective adjective is called privative; when a privative adjective is used with a noun, the effect is that
the adjective–noun combination is not a subset of the noun. Adjectives like fake and ex- are privative because a fake diamond
is not actually a diamond, and a person’s ex-husband is not their husband. The examples in (39) show that discourse -ass
associates can be privative adjectives.
(39)
Privative associates
a. (You) fake ass Suge Knight!19
b. You aint never been in anything that requires ethics you fake ass gangbanger.20
c. We don’t want your fake ass Timberlands.21
d. Gwen’s old ass boyfriend called last night. (old ¼ ex- or former)
2.2. Non-adjectival associates
Discourse -ass is productive with a noun in the associate position. Although the range of associate nouns appears to be
somewhat restricted, those nouns that do occur as associates occur with some frequency; in fact, Spears (1998: 234) discusses
discourse -ass as primarily involving a noun as the associate of the a-word. In (40), example (40a) is repeated from (13) above.
(40)
Nominal associates
a. We some punk ass parents, see. (Lee, 2000, Bernie Mac)
b. I’m a boss ass bitch. (PTAF, 2012: “Boss Ass Bitch”)22
Spears (1998) mentions that the construction is especially productive with the nouns fool, bitch and ho.
Finally, the associate in a discourse -ass construction can be a participle (41) or phrase (42), as shown below.
17
18
19
20
21
22
23
24
(41)
Participial associates
a. I got a little year older, now I’m a grown ass man. (Lee, 2000, Cedric the Entertainer)
b. Big mama .[had] those swollen ass ankles. (Lee, 2000, Bernie Mac)
c. your lying cheating ass boyfriend with that tramp from the other side of town!23
d. She’s a triflin ass woman. (Spears, 1998: 237)
e. Stinkin ass rappers make me sick. (Lee, 2000, Cedric the Entertainer)
f. My cousin Kendra and her cheating ass boyfriend.24
(42)
Phrasal associates
a. She got a fucked up ass hairdo. (Lee, 2000, Bernie Mac)
https://twitter.com/Diva2dn/status/416269700704632832.
https://twitter.com/–sheishunnie/status/261214522663112704.
Quote from the film Next Friday (2000).
http://www.defsounds.com/news/Camron_apologizes_for_snitching_comments.
http://www.hiphopgame.com/forums/showthread.php?t¼29630.
See also Nicki Minaj (2013).
http://www.nappyafro.com/Top5MusicVideos.html.
http://www.bebo.com/PhotoAlbums.jsp?MemberId¼326136266.
20
P. Irwin / Language Sciences 50 (2015) 12–29
b. Look at all them lame, ain’t got no rap ass muthafuckas.25
c. All you no money having ass wanna be rappers mad cuz he better than y’all26
d. I got .cornbread eatin ass from the city bitches (Trick Daddy, 2001: “99 Problems”)
Harley (2009) discusses phrasal compounds such as (43). The phrasal associates that are productive in discourse -ass are
perhaps best analyzed along the lines Harley suggests, through zero-derivation.
(43)
These aren’t your standard stuff-blowing-up effects. (Harley, 2009: 204)
Harley notes in addition that such phrases need not be part of a compound, since derivational morphemes can attach to them
(Harley, 2009: 223). Harley gives examples such as those in (44), where -ish and -y attach to a phrase that has first been
categorized as a nominal, before affixation by -ish and -y.
(44)
Derivational morphemes attaching to phrasal compounds (Harley, 2009)
a. feeling a bit rainy-day-ish
b. a bit ‘don’t bother’-y
On the analysis proposed here, -ass has been grammaticalized in AAE into a derivational affix (little-a head), like -ish and -y.
This analysis predicts that the modifier in discourse -ass should be able to undergo further affixation, just as rainy-day-ishness can be formed from rainy-day-ish. This is indeed the case with discourse -ass, at least to some degree.
The first use of the nominal bitchassness is attributed to the celebrity P. Diddy in season 4 (2009) of his reality television
show Making the Band (45).
(45)
Recently, especially ever since the recession, there has been an overwhelming amount of new cases of
bitchassness.27
Diddy uses the word to describe behavior that might be described in mainstream English as involving peevishness and lack of
responsibility, among other qualities.28 Popular culture jumped on this coinage, to the extent that t-shirts with the phrase
were produced and sold with the phrase “no bitchassness.”
Although Diddy’s phrase might appear to be a joke rather than the productive use of a grammatical morpheme, Google
search results can be found for similar words based on common discourse -ass associates, such as clownassness, foolassness,
hoassness and pimpassness (with only the last term having a positive connotation). Other productive affixes can also attach to
-ass, as seen Internet examples like grownass-ity, dumbassity, and punkassnessdall of which have Usrban Dictionary entries
and spelled as shown here; punkassness occurs as a Twitter hashtag. Future research will tell whether derivational affixation
on top of discourse -ass will continue to be productive.
2.3. Iterative modification and discourse -ass
Just as attributive modifiers can be iterated to modify a head N (as in the big red ball), the associate þ a-word constituent
can be iterated to modify a single head N. In other words, a single head N can have numerous discourse -ass modifiers (46):
(46)
Iterating the associate þ a-word constituent29
a. the dumb ass wack ass actor should have still been using a condom.
b. just another lame ass sad ass excuse of a man
c. a gold-diggin, triflin ass, no-talent having, pineapple-headed ass hoe
The examples in (46) (as well as (42c)) show that the associate þ a-word function as a unit in being able to be interspersed
among adjectives and other modifiers that may or may not involve an a-word. The iterability of discourse -ass is related to its
properties as an expressive, as we will see, as well as to what Spears (1998) calls the “rhythmic utility” of discourse -ass
(Spears, 1998: 237–238), discussed in x3.2.
2.4. What is the a-word?
We now turn to the status of the a-word. Determining the contribution of this element gets us close to the heart of
discourse -ass. This part of the construction has two obvious but important properties: outside of the discourse -ass construction, ass refers literally to a body part, the buttocks, and it is generally considered to be a swear word. In its literal usage,
the word patterns like a garden variety nominal (47). We will see that determining the part of speech of the a-word in
discourse -ass is not a simple task.
25
26
27
28
29
Arthur Spears, p.c.
http://www.hiphopmusicdotcom.com/lil-wayne-disses-his-dj-in-front-of-the-crowd.html.
http://www.complex.com/music/2011/11/the-10-best-diddy-rants/diddy-laments-recent-outbreaks-of-bitchassness.
See “Diddy Speaks on the Recent Outbreak of BitchAssNess” https://www.youtube.com/watch?v¼Yg5W_cHz50g.
All examples from: http://www.thelifefiles.com/eddie-murphy-mel-b-tricked-me-into-having-baby/.
P. Irwin / Language Sciences 50 (2015) 12–29
(47)
21
Girls tell me i have a cute ass30
In discourse -ass, there appears to be very little contribution of the literal meaning of the word. One way we see this is in the
fact that discourse -ass allows for almost no a-word substitutions. Collins et al. (2008) report that in the camouflage construction they discuss, butt, behind, or tail, words that in their literal usage denote the body part, are acceptable in situations
where swearing is not appropriate (Collins et al., 2008: 10).
(48)
I’m gonna fire his sorry ass/behind/butt/tail. (Collins et al., 2008, 10).
With discourse -ass, the a-word is almost categorically restricted to ass. Spears (1998) reports that discourse -ass does not
allow any substitute word. The informants consulted for this paper accepted some alternate a-words, but in general they
judged them as either degraded or unacceptable. They accepted behind and butt (49), but rejected tail and fanny (50).
(49)
Acceptable alternate a-words
a. ? Gwen’s stupid behind husband forgot the beer again.
b. ? Gwen’s stupid butt husband forgot the beer again.
(50)
Unacceptable alternate a-words
a. *Gwen’s stupid tail husband forgot the beer again.
b. *Gwen’s stupid fanny husband forgot the beer again.
Note that the judgments in (49) are contrary to those reported by Spears (1998); this may be due to dialect variation.31
In other words, ass of discourse -ass is special in having been “semantically bleached” (Spears, 1998: 236). The other awords have not undergone such a semantic bleaching process, so they are not acceptable in discourse -ass. Although ass is
bleached of its literal meaning, however, it still retains the force of a moderate swear word. Before discussing these aspects of
the contribution of the a-word, we must first describe its syntactic properties.
2.4.1. Syntactic properties of the a-word
Although the literal use of ass patterns like an ordinary N, the properties of the word in discourse -ass make determining
its part of speech (grammatical category) less clear-cut. The a-word resists all inflection: it cannot be made plural, and it
cannot be a possessor. For example, even in contexts where the a-word is part of a structure that modifies a plural head N (51),
it cannot be made plural.
(51)
The a-word can be made plural
a. *There are some crazy-asses people who live on my street.
b. *You should throw away them ugly-asses clothes.
c. *Our stupid-behinds boyfriends forgot the beer again.
d. *I’m gonna sue those rich-asses people!
The a-word also cannot be a possessor, as shown in (52).
(52)
a. *Jamie’s ugly ass’s car was broken.
b. Jamie’s ugly ass car was broken.
The head N that is modified by the associate þ a-word can be a possessor, however (53a). As shown in (53b), the possessive
morpheme attaches to the whole head N constituent.
(53)
a. Jamie’s ugly ass car’s fender was broken.
b. [ [ jamie ] [ ’s ugly-ass car ] ] ’s fender
These properties make the a-word pattern like a light or perhaps bare noun, such as home (as discussed in Collins, 2007: 10,13;
see also Kayne, 2005: 256ff). Lexical items that are in “light” categories may be semantically bleached, and they often have
both functional properties and lexical properties.
2.5. Summary
The previous subsections presented the syntactic distribution of discourse -ass and, within the construction, the properties of the associate and the a-word. We have seen that the associate þ a-word constituent is attributive: it must occur
before the head N, and it cannot function as a predicative complement. We have also seen that the associate may be an
adjectivedwith few restrictions on semantic classda noundwith more restrictionsdor a participle or phrase. Nominal
associates cannot be inflected (e.g., *punks-ass boys), and adjectival associates cannot be modified. Adjectival associates
cannot be modified with very, for example, and they cannot be made comparative or superlative (e.g., *cutest-ass boy).
30
31
https://www.facebook.com/pages/Girls-tell-me-i-have-a-cute-ass-but-when-i-tell-them-they-do-they-get-mad/306525876979.
Arthur Spears (p.c.) notes that rhythmic/metrical constraints may also be affecting judgments on these data.
22
P. Irwin / Language Sciences 50 (2015) 12–29
2.5.1. Arguments for and against a compounding analysis
Spears (1998, 236) discusses the possibility that discourse -ass involves “a type of compound,” but he also notes several
ways in which the modifier constituent of discourse -ass lacks certain properties of English compounds. In this section we will
consider the ways in which the modifier appears to form a compound, but we will see that a compounding analysis is ultimately not the best route to pursue.
Why might we think that the modifier forms a compound? The modifier consists of two lexical items, the associate and the
a-word, and these lexical items seem to form a word that is adjectival in nature; one reason for this is the fact that the whole
modifier constituent can be made comparative or superlative. In addition to the intuitive appeal of a compounding analysis, it
would also account for two important properties of the modifier.
The first argument in favor of a compounding analysis is that it would explain why the associate cannot be inflected.
English compounds famously do not allow inflection of the non-head (left) word in a compound. This is illustrated in (54) for
Adj–Adj compound bitter sweet, the synthetic compound truck driver, and the N–N compound nail salon.
(54)
Inflection on the non-head word in English compounds is ungrammatical
a. the bitter sweet memory
b. *the bitterest sweet memory
c. a truck driver
d. *a trucks driver
e. a nail salon
f. *a nails salon
As shown in x2.1.1, the associate in a discourse -ass modifier cannot be inflected. This is illustrated in (33), repeated below as
(55).
(55)
No
a.
b.
c.
inflection on the associate
This my cute ass friend.
*This my cuter ass friend.
*This my cutest ass friend.
Another property of compounding is simply the fact that a compound word acts like a complete unit. For example, more
attaches to the whole Adj–Adj compound in (56a), and the plural morphology -s in (56b) attaches to the whole word, not just
the second word (57).
(56)
a. the most bitter sweet memory
b. several truck drivers
(57)
several [[truck driver]s]
In the case of discourse -ass, we see that the associate þ a-word form a unit; the superlative morphology in (58) attaches to
the associate þ a-word as a unitdnot to just the associate.
(58)
I have the most fool ass sense of humor.32
In other words, the structural relationship of the discourse -ass modifier to “sense of humor” in (58) is as shown in (59a), and
not as in (59b).
(59)
a. the [most [fool ass] ] sense of humor
b. *the [ [most fool] ass] sense of humor
For comparison, note that in mainstream English, the superlative most happily modifies the overtly adjectival foolish, but not
fool (60). These data provide additional support for an analysis in which -ass is a functional, adjectival head.
(60)
a. the [most foolish] sense of humor
b. *the [most fool] sense of humor
Finally, as discussed in x1.3, the discourse -ass modifier has the stress pattern associated with English compounds, where
primary stress goes on the first element, and secondary stress goes on the second element.
Despite the fact that the discourse -ass has these properties of English compounds, there are numerous challenges to a
compounding analysis. We saw in x1.3 that the stress pattern of English compounds is also shared by words with derivational
morphology like green-ish and crisp-y, supporting an analysis in which -ass is a functional morpheme. Such an analysis is
strengthened by the fact that the word ass is semantically bleached to the point of hardly being paraphrasable; this semantic
property is discussed further in x3.1.4.
A more complex challenge to a compounding analysis concerns the basic question of what part of speech the a-word is. In
its literal usage, ass is a N. If we assume that the a-word in discourse -ass is a N, then, if we assume the a-word þ associate
32
http://twitpic.com/photos/MzDemeanor?page¼3.
P. Irwin / Language Sciences 50 (2015) 12–29
23
functions as an adjective, then we get an undesirable result: A phrase like proud ass would be an Adjective–Noun compound
that functions as an Adjective. In English, however, the right-hand element in compound words almost always determines the
category of the compound. This has been generalized as the Right-hand Head Rule (RHL), which goes back at least to Williams
(1981).
(61)
[ blackAdj birdN ]N
Semantically, it has long been known that the right-hand element in English compounds determines the type of the result, in
what has been called an “is-a” relationship (Allen, 1978). For example, a blackbird is a type of bird, not a type of black; a truck
driver is a type of driver, not a type of truck, and so on.
English does have left-headed Adj–N compounds, however (62), so it is possible that discourse -ass is one of them.
(62)
Possible left-headed compounds: Adj þ N / Adj
a. a last-minute errand
b. a deep-sea adventure
But the fact remains that left-headed compounding is fairly unproductive in English, a language with otherwise abundant
(right-headed) compounding. Even the examples in (62) are suspect as straightforward adjectival Adj–N compounds, since
deep-sea in (62b), for example, has a shaky status as an adjective (*a very deep-sea adventure).
The foregoing syntactic facts lead us away from a simple Adj–N compound analysis. Let us now consider the possibility
that the associate þ a-word is a compound, but that ass is not a N as it is in its literal usage, but is an Adjective. This analysis is
schematized in (63).
(63)
a [ proudAdj assAdj ]Adj fan
Adjective–Adjective compounding is somewhat productive in English (blue-green, bittersweet), but this analysis faces its own
problems. The first problem is that if the associate þ a-word constituent functioned as an adjective, we would expect it to be
modifiable with adverbs such as very and awfully, as we see with the garden-variety Adj–Adj compounds in (64).
(64)
a. awfully bittersweet memories
b. very blue-green eyes
Although modification with most may be possible in some cases, as in (58) above, discourse -ass in general resists this kind of
adverbial modification (65).
(65)
a. *We’re some very punk-ass parents.
b. *Jamie’s wearing an awfully raggedy-ass sweater.
The second set of problems with the Adj–Adj compounding analysis concerns the a-word. The word ass does not independently pattern as an adjective in any variety of English that I know of (66).
(66)
a. *a very ass game
b. *a really ass book
Second, ass is semantically bleached, as noted repeatedly above. Compounding, on the other hand, involves a semantic
contribution from all of its parts. Our basic understanding of compounding is that it involves “two independently meaningful
roots” that are “directly combined to form a new, complex word” (Harley, 2006: 99). In its literal meaning, the word ass
functions as a syntactically-contentful root, but in discourse -ass it does not contribute any of the descriptive meaning
associated with that root.
Although this is as far as we will go with the syntactic analysis of discourse -ass, it seems clear that the associate þ a-word
is not a compound. As Spears (1998) points out, ass in the AAC likely arose historically from “anatomical ass”, but in discourse
-ass, the word ass has been grammaticalized Spears (1998: 236)dit has become a functional morpheme.
The semantic bleaching of ass and the challenges this creates for determining the syntax of the discourse -ass construction
leads directly to the question of what the discourse -ass construction means. Following Spears (1998), I have suggested that
ass has been grammaticalized into a functional, adjectivizing, morpheme, like -ish in fool-ish, and -y in crisp-y. But intuitively,
ass contributes more to the meaning of its associate than -y does to its root, as shown in (67) (brackets added). But what is the
difference?
(67)
a. Listening to these [crisp ass] beats by D. Harliquinn33
b. Listening to these [crispy] beats by D. Harliquinn
We turn to these questions next.
33
https://www.facebook.com/GigiSmithMusic/posts/406834026054089.
24
P. Irwin / Language Sciences 50 (2015) 12–29
3. Meaning and discourse -ass: What is the difference between an ugly cat and an ugly ass cat?
3.1. Expressive properties
Intuitively, discourse -ass appears to be expressive: it requires a word that is considered inappropriate if not vulgar in
polite speech settings, and it occurs very often as part of African American communicative practices that have an element of
performance in them. Discourse -ass is also expressive in a more technical sense, as first observed by Spears (1998):
The meaning of -ass in AWs is to be found on the level of discourse and expressive meaning. That is, what -ass in AWs
communicates is invariably something about the communicative situation. (Spears, 1998: 236)
In other words, the difference between an ugly cat and an ugly ass cat has nothing to do with the cat, or with the property of
ugliness.
Spears writes that discourse -ass communicates that the utterance is in “uncensored mode”, a mode of discourse that may
involve expressions that may be considered obscene but that are “used in an almost or completely evaluatively neutral way”
(Spears, 1998: 236). Censored discourse is associated with speech situations like church and school (Spears, 1998: 232), and
uncensored discourse is associated with situations where stigmatized speech is permitted.
It comes as no surprise, therefore, that much of the data for this paper comes from stand-up comedy acts, which in both
black and white American cultures involve a discourse situation where “uncensored” speech, in the everyday sense of the
term, is the norm. We might think of “uncensored mode” in Spears’s sense as being a way of speaking that falls under the
broader principle of “directness,” which Spears discusses in subsequent work on African American communicative practices
(Spears, 2001, 2007).34
Recent work on the meaning of expressive elements in natural language has identified several descriptive diagnostics for
expressivity. Potts (2007b) proposes six characteristics of expressive content. These characteristics are particularly revealing
about how expressive content differs from descriptive content, which we will see in examining discourse -ass. We will see
that discourse -ass has the properties proposed in Potts (2007b), thus supporting Spears’s early observations that the construction contributes meaning at the level of the discourse context rather than at the truth-conditional (semantic) level.
Potts (2007b) proposes that expressive content has the following six properties: independence, non-displaceability,
perspective independence, descriptive ineffability, immediacy, and repeatability. The following sections present each of
these properties in turn and discuss the ways in which discourse -ass has (or does not) have them.
3.1.1. Independence
The INDEPENDENCE of expressive meaning refers to the idea that expressives “contribute a dimension of meaning that is
separate from regular descriptive content” (Potts, 2007b: 166). In its simplest form, the property of independence is seen in
the fact that an expressive does not change the truth conditions of the sentence it occurs in. For example, the sentences in
(68a) and (68b) with and without the expressive damn have the same truth conditions; in other words, they have the same
descriptive content.
(68)
a. I took out the damn trash this morning.
b. I took out the trash this morning.
An expressive contributes meaning not in the semantic/truth-conditional domain (also called the descriptive content), but in
another dimension of meaning, what Potts and others call the “expressive domain”. This generalization holds for discourse
-ass.
Consider the sentences in (69). These sentences both have the descriptive content in the gloss for (69b). The presence or
absence of -ass does not change this descriptive content.35
(69)
a. I don’t know why you got that pink ass shirt on. (Binkow, 2007, Mo’Nique)
b. I don’t know why you got that pink shirt on.
z ‘I don’t know why you’re wearing that pink shirt’
It is clear that (69a) and (69b) have different meanings, but this meaning difference is not truth-conditional.
The property of independence raises the question of what semantic contribution the discourse -ass construction in (69a)
really makes. One claim in the literature is that ass serves to give an expression “extra emphasis” (Smitherman, 1994, 2000:
56–57). Spears (1998) argues that ass does not always contribute emphasis, though he writes that it may “in specific utterances have the force of an intensifier” (Spears, 1998: 236). As shown in (70), the modifier in discourse -ass can be intensified,
at least with damn.
(70)
He wears the same damn ugly ass tie every day!
34
Directness is most obviously exemplified in African American communicative practices such as playing the dozens, snapping, cussin’ out, reading
people, etc. See Spears (2001, 2007) and the copious literature on these practices for more discussion.
35
This point holds despite the fact that this example occurs as part of a factive predicate.
P. Irwin / Language Sciences 50 (2015) 12–29
25
Potts (2007b,a) and others argue that expressives contribute to meaning by modifying the speech situation. Potts
proposes a specific formal mechanism according to which the expressive, as its own semantic type, composes with
descriptive (truth-conditional) meaning and the speech context to return the descriptive meaning with the speech context
changed (Potts, 2007b, 187).36 This implementation captures the intuition that expressives do not operate at the level of
truth conditions, but they do affect the context in which an utterance is evaluated. In other words, the meaning of an
utterance with an expressive is interpreted with respect to a new context, one that is brought about by the use of the
expressive.
The formal denotation of Potts’s analysis is given in (71), for interested readers, where is an operator that Potts introduces
to indicate the composition of expressive meaning with descriptive meaning. In (71), s is a descriptive type, 3 is an expressive
type, <s, 3 > is an expressive type, and c is the context (Potts, 2007b, 183).
(71)
The effect of an expressive on the context (Potts, 2007a: 258)37
Where a is of type <
s, 3c> and b is of type s,
c0
c0
c
EaF ðEbF ÞðcÞ
EaF EbF ¼ EbF
We will not pursue the analysis of discourse -ass with respect to these technical details, but it should be clear that Potts’s
analysis formally implements Spears’s early observations on discourse -ass and other “obscene” language, which is that such
language does not affect truth conditions but does bring about a particular speech context.
The data in (69a) also raises an important point about the meaning of discourse -ass: the discourse -ass construction does
not in itself imply a positive or negative evaluation of the head N. In (69a), the evaluation is neutral. This property separates
discourse -ass from other expressives such as slurs, which until recently (see Croom, 2014 for some discussion) were seen as
always contributing a negative evaluation of what they modify. In discourse -ass, any evaluative force comes from the
associate, not from the word ass (Spears, 1998: 237–238). And indeed, examples like boss-ass bitch (40b) are used with a
wholly positive evaluation. As Spears points out, it is a common misconception that discourse -ass can only be used for
negative evaluation. The idea that discourse -ass can be used only for negative evaluation may come from the fact that when
white speakers of mainstream English use the discourse -ass (or a construction like it), it can only be used for negative
evaluation.
The examples in (72) show that discourse -ass can be used in utterances that express the speaker’s positive evaluation of
the head N.
(72)
Positive evaluation associates in discourse -ass
a. Ewing got some dope ass shoes38
b. My cute ass cousin.39
c. You’re all laughing, you’re having a good ass time. (Binkow, 2007, Mo’Nique)
Discourse -ass may be among the “purest” of expressives in not contributing the speaker’s (positive or negative) evaluation
of the head N, and it does not even serve to focus or otherwise emphasize the meaning of the associateda pink ass shirt is not
any more pink or shirt-like than a pink shirt. In terms of meaning, the difference between an utterance with and without
discourse -ass purely concerns the speech context, where discourse -ass marks the context as performative (in Spears’s sense)
and direct.
3.1.2. Non-displaceability
NON-DISPLACEABILITY is the property that expressives have of attaching themselves to the current utterance situation, even if
the content of the utterance concerns a past situation or attitude (Potts, 2007b: 169). To illustrate this with our example from
(68), consider the use of damn in mainstream English (73).
(73)
John told me that he took out the damn trash this morning.
We cannot interpret damn in this sentence as commenting on John’s attitude toward the trash; damn reaches out of its
embedded context and commits the speaker to a particular attitude (in this case negative) of the trash or the event of trash
removal. The only way the expressive in (73) could possibly report on the embedded situation is through direct quotation,
which would be marked in an utterance like (73) with special intonation on damn trash.
Discourse -ass, in contrast to an expressive like damn, does not contribute any evaluative meaning, so it is harder to put it
into a test like (73) or those that Potts gives with bastard. The only evaluative force in the discourse -ass construction comes
from the associate itself. In order to test non-displaceability, consider (74).
36
The context consists of: speaker/agent, time, world, judge, and a set of expressive indices (Potts, 2007b: 184).
This denotation comes from Potts (2007a), which includes a revision of the denotation given in (Potts, 2007b: 187) in response to the replies to the
paper.
38
http://theshoegame.com/articles/dallas-penns-drop-10-what-not-to-wear-april-2014.html.
39
http://s195.photobucket.com/user/123 JROC 123/media/ALAYSIAISON.jpg.html.
37
26
P. Irwin / Language Sciences 50 (2015) 12–29
(74)
Non-displaceability and embedding
a. Jamie told me he got some dope sneakers, but I saw them and they’re ugly!
b. ??Jamie told me he got some dope-ass sneakers, but I saw them and they’re ugly!
The relevant contrast in (74) is not about the speaker’s disagreement with the content in the reported speech, but about
which speech situation discourse -ass in (74b) attaches to. Because discourse -ass wants to attach to the immediate speech
situation, the embedding in (74b) sounds strange if dope-ass is meant to describe the speaker’s attitude toward the sneakers.
Note that this infelicity holds whether the speaker agrees or disagrees with the reported evaluation (75).
(75)
??Jamie said he got some dope ass sneakers, and I agree they’re fresh.
Discourse -ass simply sounds odd in reported speech, and this is because it contributes its meaning to the current speech
situation. Perhaps most interestingly, the non-displaceability of the a-word has the effect of making the evaluative force of the
associate reach out of its embedded context. In other words, part of the oddness of (74b) and (75) is the fact that because dope
is an associate of ass, it also wants to describe the speaker’s attitude rather than that of the matrix subject.
3.1.3. Perspective dependence
PERSPECTIVE DEPENDENCE refers to the idea that expressives encode an individual’s attitude toward some “feature of the current
state of affairs” (Potts, 2007b: 173). This may often be emotional, broadly construed: prototypical uses of damn reflect anger or
frustration; pronouns in Korean can express the speaker’s humbleness toward the interlocutor; honorifics express the
speaker’s respect toward a third party (Hwang, 1990, inter alia).
As we have seen, discourse -ass does not necessarily reflect heightened emotion, and in this way it may be more akin to the
“expressive” content of speech styles in a language like Korean: they say something about the speech context. The property of
PERSPECTIVE DEPENDENCE has been much discussed with respect to the notion of context-shifting, as Potts (2007b) discusses (see
Schlenker, 2003 and subsequent responses). These matters take us too far afield for this overview, and for our purposes,
PERSPECTIVE DEPENDENCE may be seen as linked very closely to NON-DISPLACEABILITY, as discussed above.
3.1.4. Descriptive ineffability
Discourse -ass pointedly manifests the expressive property of DESCRIPTIVE INEFFABILITY. Simply put, descriptive ineffability
refers to the fact that it’s hard to paraphrase the meaning of an expressive. Even a slur like bastard is difficult to paraphrase in
its non-literal use, since it can have affectionate uses, as many have pointed out (Croom, 2014; Potts, 2007b: 176–177).
Terms like bastard or bitch can be used either affectionately or disparagingly40, and for most examples it’s clear which use
the speaker intends. As we have seen, discourse -ass does not have an effect on the evaluative scale. Consider Mo’nique’s
comment from (69a), given below as (76).
(76)
I don’t know why you got that pink ass shirt on. (Binkow, 2007, Mo’Nique)
It is clear from the context that ass in (76) does not express a positive or a negative evaluation of the pink shirt, nor does it
pick out a degree of pinkness. Examples like (76) argue against a possible analysis of the a-word as a degree marker.41 Because
discourse -ass does not increase or decrease an affectionate or disparaging evaluation, it is hard to characterize its meaning in
terms of the expressive indices proposed in Potts (2007b) and Potts and Kawahara (2004).
3.1.5. Immediacy
Potts (2007b: 180) points out that expressives are like performatives in that “the act of uttering an expressive morpheme is
sufficient for conveying its content.” For slurs and epithets, for example, the act of uttering the phrase does the expressive
work. I have argued above that discourse -ass does not reflect either positive or negative evaluation, but, following Spears
(1998), that it marks the speech context as direct and uncensored. The marking of a context as direct/uncensored, in an
African American speech context, can, in addition, create solidarity between speaker and interlocutor.
We see a particularly effective use of context modification in the first three lines of Steve Harvey’s stand-up comedy
monologue from Lee (2000). These lines are given as three linguistic examples in (77)dnote that only (77c) contains
discourse -ass (highlighted).
(77)
40
Steve Harvey: opening monologue (Lee, 2000)
a. How y’all doin with y’all country ass? [audience laughs]
b. You know your ass is country! Don’t be looking around!
c. I’m country my own damn self, that’s how I know country ass people when I see ’em!
For numerous examples of bitch used affectionately, see Binkow (2007).
The example in (76) raises the possibility that ass may serve as a type of focus marker; the fact that the shirt is pink is highlighted by Mo’nique in this
context, as a way of teasing the interlocutor for wearing a color associated with femininity. As we have seen in many examples here, however, not all
occurrences of discourse -ass can be reduced to focus.
41
P. Irwin / Language Sciences 50 (2015) 12–29
27
These lines use “uncensored” speech to mark the speech context as direct, performative, and thereby African American.
Syntactically, Harvey uses several patterns specific to African American English: the “resumptive-with” construction with y’all
country ass in (77a) (for discussion of this construction, see Collins et al. 2008: 51–54); in (77a), the lack of overt genitive
marking on the possessor y’all (Green, 2002: 102–103); the camouflage constructions y’all country ass in (77a) and your ass in
(77b).
Harvey’s use of ass in the camouflage construction and as discourse -ass is particularly effective in changing the speech
context. This is in part because Harvey needs to effect such a change, and he needs to do so without being offensive to fans
who support his (recent, at the time) self-presentation as a Christian.42 A significant portion of his routine discusses his
attempts to reduce the “cussin” in his act (“I’m not a perfect Christian,” Lee, 2000). But at the same time, a standup routine is a
context that asks for uncensored speech. Therefore, the use of discourse -ass as a way to change the context to one of uncensored speech is important in preparing the audience for his performance.
3.1.6. Repeatability
The last property of expressive content, REPEATABILITY, is particularly interesting with respect to discourse -ass. Repeatability
refers to the fact that for expressives, “repetition leads to strengthening rather than redundancy” (Potts, 2007b: 182). Potts
gives the example in (78), where the expressive damn is repeated but does not have the effect of providing redundant
information.
(78)
REPEATABILITY with damn (Potts, 2007b, 182)
a. Damn, I left my keys in the car.
b. Damn, I left my damn keys in the car.
c. Damn, I left my damn keys in the damn car.
Discourse -ass clearly has the property that repeated uses only strengthen its effect; the examples from Steve Harvey’s
monologue (77) illustrate this, assuming that the ass in the camouflage construction is semantically also a type of expressive.
As discussed in x2.3, discourse -ass can be iterated throughout a sequence of attributive modifiers. This was shown in (46), as
well as below in (79).
(79)
a. All these lame ass, punk ass, bitch ass niggas/hoes wanna go for bad and know they ain’t gone bust a fruit
in a fruit fight.43
b. These lame ass punk ass weak ass sorry ass hoe ass pussy ass can’t-rap-for-shit-ass niggaz is scared to
battle!44
The fact that discourse -ass has the expressive property of repeatability is linked to its prosodic effect, which is one of its most
important features from a speech usage point of view. Spears (1998) observes that “AWs are often used by the speaker simply
because of their rhythmic utility,” when an important feature of the speech situation is “expressive ingenuity and social
effect” (Spears, 1998: 237–238). What is the rhythmic utility of discourse -ass? The following section suggests an answer to
this question.
3.2. The rhythmic utility of discourse -ass
The previous sections have discussed how the meaning of the a-word in discourse -ass exists purely on the expressive
level: in Spears’s words, it serves the role of “marking a discourse as direct (Spears, 2001) and/or heightening the poetic and
performance character of what is being said” (Spears, 2007: 231). This section will address the way in which discourse -ass has
rhythmic effects that are related to English metrical structure. In particular, I discuss the way in which discourse -ass has the
rhythmic effect of reducing what is called “stress clash” (Hammond, 1988).45
A stress clash occurs when two stressed syllables are next to each other. Crosslinguistically, languages vary in their
tolerance for stress clashes and in the means they can employ to reduce them. Somewhat famously, English has a mechanism
that has been called the Rhythm Rule (see Liberman and Prince (1977), who cite Kiparsky (1966) for an early discussion).
Informally put, this mechanism allows English to shift stress backward to eliminate a stress clash. For example, in isolation,
the word thirteen has main stress on the second syllable (80a). But when the first syllable of a word following thirteen is
stressed, the stress in thirteen can be shifted backward, to the first syllable, as shown in (80b).
(80)
a. thirtéen
b. thírteen bóoks
42
The first lines of his act, immediately preceding those in (77), are: “Before I get started, I never start anything without first off saying that God is
everything to me.” (Lee, 2000).
43
https://twitter.com/AK97/status/207182558927134720.
44
http://www.dopetracks.com/forums/11/topics/1654 (punctuation and capitalization modified).
45
See Siddiqi (2011) for an interesting and very different approach to the morpho-phonological constraints on discourse -ass. Siddiqi’s squib appears to
focus on uses of -ass in mainstream English rather than in AAE.
28
P. Irwin / Language Sciences 50 (2015) 12–29
Now consider the phrase some punk-ass parents. Without the a-word, this phrase would contain a stress clash, as shown in
(81).
(81)
Stress clash without discourse -ass
some púnk párents
In the case of (81), it is not possible to move the stress on punk leftward to some, since that would change the meaning. The
Rhythm Rule cannot apply in this context.
(82)
Rhythm Rule cannot apply
sóme punk párents
The insertion of ass, on the other hand, fixes stress clash and does not change the semantic meaning of the phrase. In
contexts like (83), primary stress on the associate and secondary stress on ass result in a perfectly trochaic foot, a much more
favorable metrical structure. This is shown in (83), with primary stress indicated.
(83)
Discourse -ass: no stress clash
some púnk-ass párents
In general, when the associate is monosyllabic or has final stress, the a-word results in a modifier with a trochaic foot. This
means that sequences of discourse -ass modifiers are sequences of trochees whenever the associate is monosyllabic, as it
often is.
Consider two versions of one of the examples in (46), repeated below in (84) and (85). In (84), the sequence lame ass and
sad ass has two trochaic feet.
(84)
Discourse -ass sequence with two trochaic feet
Just another láme ass, sád ass mán
From the metrical standpoint of English, the variant without the a-words is suboptimal, since it contains a series of stress
clashes (85).
(85)
No discourse -ass: Multiple stress clashes
Just another láme, sád mán
The contrast between (84) and (85) illustrates the “rhythmic utility” of discourse -ass with relatively simple examples. When
sequences of discourse -ass become longer, with phrasal associates, for example, the prosodic considerations become more
complex. The question of when a speaker would choose to insert an a-word based on rhythmic considerations also raises deep
questions about prosodic-driven variation and grammatical optionality that are beyond the scope of this paper.
4. Conclusion
The discourse -ass construction has a complex syntax, semantics and phonology. The overview in this paper has not
answered every question about the construction, but it has outlined the basic facts on which further, more detailed and
technical syntactic and semantic research can be done. Syntactically, one of the mysteries of the construction is the nature of
the word ass, which I have argued has the syntactic and prosodic properties of a functional item, despite the fact that in its
literal usage the word is nominal. Semantically, it is clear that the construction has the properties of an expressive, as defined
by Potts (2007b). Many of the semantic properties of the construction were first described in Spears (1998), who observed
that the a-word in discourse -ass has undergone semantic bleaching. For discourse -ass, one of the most prominent of its
semantic properties, in terms of Potts (2007b), is INDEPENDENCEdthe fact that an utterance with and without discourse -ass have
the same truth conditions. We have also seen that the a-word does not in itself contribute speaker evaluation of the head N;
only the associate does that. This is another property of expressive speech. Like all expressives, the a-word changes the speech
context. It changes the context to what Spears (1998) calls “uncensored mode,” a mode that plays a role in numerous African
American communicative practices.
References
Adams, Michael, 2004. Slayer Slang: a Buffy the Vampire Slayer Lexicon. Oxford Univ Press, Oxford.
Allen, Margaret R., 1978. Morphological Investigations (Doctoral dissertation). University of Connecticut.
Beavers, John, Koontz-Garboden, Andrew, 2006. A universal pronoun in English. Linguist. Inq. 37.
Binkow, Gary (Producer and Director), 2007. Mo’Nique: I coulda been your cellmate (motion picture). Salient Media (DVD), United States.
Collins, Chris, 2007. Home sweet home. NYU Work. Pap. Linguist. 1.
Collins, Chris, Moody, Simanique, Postal, Paul, 2008. An AAE camouflage construction. Language 84, 29–68.
Collins, Chris, Postal, Paul Martin, 2012. Imposters: a Study of Pronominal Agreement. MIT Press.
Croom, Adam M., 2014. The semantics of slurs: a refutation of pure expressivism. Lang. Sci. 41, 227–242.
Green, Lisa J., 2002. African American English: a Linguistic Introduction. Cambridge University Press, Cambridge, New York.
Hammond, Michael Theodore, 1988. Constraining Metrical Theory: a Modular Theory of Rhythm and Destressing. Garland.
Harley, Heidi, 2006. English Words: a Linguistic Introduction. Blackwell, Cambridge, MA.
P. Irwin / Language Sciences 50 (2015) 12–29
29
Harley, Heidi, 2009. Compounding in distributed morphology. In: Lieber, Rochelle, Stekauer,
Pavel (Eds.), The Oxford Handbook of Compounding. Oxford
Univ Press, pp. 204–228 (Online Publication Date: September 2012).
Huddleston, Rodney D., Pullum, Geoffrey K., 2002. The Cambridge Grammar of English. Cambridge University Press, Cambridge.
Hwang, Juck-Ryoon, 1990. ‘Deference’ versus ‘politeness’ in Korean speech. Int. J. Soc. Lang. 1990, 41–56.
Kamp, Hans, Partee, Barbara, 1995. Prototype theory and compositionality. Cognition 57, 129–191.
Kayne, Richard S., 2005. Silent years, silent hours. In: Kayne, Richard S. (Ed.), Movement and Silence. Oxford University Press, Oxford, pp. 241–260.
Kiparsky, Paul, 1966. Ueber den Deutschen Akzent. In: Studia Grammatica. Akademie-Verlag, Berlin.
Lee, Spike (Producer and Director), 2000. The Original Kings of Comedy (motion picture). Paramount Home Video, United States.
Liberman, Mark, Prince, Alan, 1977. On stress and linguistic rhythm. Linguist. Inq. 8, 249–336.
Lieber, Rochelle, 1992. Deconstructing Morphology. The University of Chicago Press, Chicago-London.
Minaj, Nicki, 2013. Boss ass bitch (remix).
Potts, Christopher, 2007a. The centrality of expressive indices. Theor. Linguist. 33, 255–268.
Potts, Christopher, 2007b. The expressive dimension. Theor. Linguist. 33, 165–198.
Potts, Christopher, Kawahara, Shigeto, 2004. Japanese honorifics as emotive definite descriptions. In: Young, Robert B. (Ed.), Proceedings of SALT 14, vol. 14,
pp. 253–270.
PTAF, 2012. Boss ass bitch [non-album single].
Schlenker, Philippe, 2003. A plea for monsters. Linguist. Philos. 26, 29–120.
Siddiqi, Daniel, 2011. The English intensifier ass. Snippets, 16–17.
Smitherman, Geneva, 1994, 2000. Black Talk: Words and Phrases from the Hood to the Amen Corner. Houghton Mifflin, Boston.
Spears, Arthur K., 1998. African-American language use: ideology and so-called obscenity. In: Mufwene, Salikoko S., Rickford, John R., Bailey, Guy,
Baugh, John (Eds.), African-American English: Structure, History, and Use. Routledge, London, pp. 226–250.
Spears, Arthur K., 2001. Directness in the use of African-American English. In: Lanehart, Sonja L. (Ed.), Sociocultural and Historical Contexts of AfricanAmerican English. John Benjamins, Amsterdam & Philadelphia, pp. 239–259 (Chapter 9).
Spears, Arthur K., 2007. African American communicative practices: performativity, semantic license and augmentation. In: Samy Alim, H., Baugh, John
(Eds.), Talkin Black Talk. Teachers College Press, New York, pp. 100–111.
Trick Daddy, 2001. “99 Problems.” Thugs are Us. CD. Slip-n-Slide Records. Lyrics from: http://www.azlyrics.com/lyrics/trickdaddy/99problems.html.
Williams, Edwin S., 1981. On the notions ‘lexically related’ and ‘head of a word’. Linguist. Inq. 12, 245–274.