Topic: Tape a conversation with some of your friends talking about race, or gender. Examine the talk for indexicality and negotiations regarding race.

Haskell: Unconscious linguistic referents to race 59

Discourse & Society Copyright © 2009 SAGE Publications

(Los Angeles, London, New Delhi, Singapore and Washington DC)

www.sagepublications.com Vol 20(1): 59–84

10.1177/0957926508097095

Unconscious linguistic referents to race: analysis and methodological frameworks

R O B E R T E . H A S K E L L U N I V E R S I T Y O F N E W E N G L A N D , U S A

A B S T R A C T Recent years have seen considerable development in methodological designs for accessing and eliciting unconscious cognitive schemata in response to social stimuli, including race. One design is experimental and involves the priming and automatic activation of schemata. Another design is a specifically developed psycho-linguistic and logico- mathematic method for recognizing, analyzing, and validating unconsciously expressed meaning in verbal narratives, referred to as sub-literal (SubLit). Unconscious references to race found in verbal discourse from both laboratory and everyday settings, including the mass media, are illustrated and analyzed utilizing these two complementary methodological designs. Along with suggested procedures, an initial classification scheme for attributions of racial stereotypes and prejudice are presented. Given the historical and current state of race relations both nationally and globally, the paper has significant social implications.

K E Y W O R D S : discourse analysis, linguistic schemata, prejudice, racism, slips-of-the-tongue, social cognition, stereotypes, unconscious

That negative stereotypes about race remain currently active seems a truism to some; to others, stereotypes about race are no longer thought to be significantly present in social discourse. Clearly, overtly expressing negative attitudes about race has become socially unacceptable. There appears to be only a small group of people who would now openly admit to being racist, or to holding negative stereotypes about race (see Van Dijk, 1992). Nevertheless, race relations remain both complex and increasingly subtle (see Gresson, 1982, 1995).

Not infrequently, the mass media report unconscious ‘racist’ slips-of-the- tongue, ostensibly pointing to hidden negative stereotypic schemata. Recently, a radio talk show host, in voicing the possibility of Secretary of State Condoleezza Rice becoming football commissioner (as she has indicated she might like),

A R T I C L E

60 Discourse & Society 20(1)

consciously meant to say that being African-American, her landing the job would be a big coup, but ‘slipped’ and said she would be a ‘big coon.’1 In another recent example, a high school athletic director is reported to have ‘slipped’ and referred to Martin Luther King Day as ‘Martin Luther Coon Day.’2 Similarly, a weather forecaster made the same ‘slip.’3 Numerous other examples are reported in the mass media and across the internet.

Perhaps somewhat less offensive, in the immediate context of a remark by Condoleezza Rice that racism was a kind of ‘birth defect’, during a promo for his upcoming program the CNN anchor Lou Dobbs, in railing against ‘politically correct’ responses when talking about race, maintained that most Americans do not have a problem talking about race, only about being called racist when they do.4 Continuing, Dobbs consciously intended to say, ‘These ridiculous politicians . . .,’ but made a verbal ‘slip,’ saying, ‘These cotto . . .,’ the full meaning of which would have been ‘These cotton picking politicians.’ But he caught and corrected himself in mid-syllable, ending with ‘These ridiculous politicians . . .’

‘Slips’ about race are sometimes made by people who might not be suspected of harboring negative stereotypes about race. One widely published incident involved the famous former sportscaster, Howard Cosell. During a 1983 play-by- play description of the African-American Washington Redskins wide receiver, Alvin Garrett, Cosell described Garrett running down the field by saying, ‘Look at that little monkey run.’5 The immediate, widely publicized and understandable reaction was to see this remark as an unconscious slip to the racist belief that African-Americans descended from monkeys and apes.

In 1999, a slip widely reported in the mass media about the use of the word niggardly by a white man in a meeting of mostly African-Americans was imme- diately interpreted as a racist slip by many members of both the White and African-American communities. Since the word has no linguistic connection to the ‘N-word,’ the NAACP immediately disavowed that any racism was implied by the use of the word (New York Times, 1999). A closer linguistic analysis, however, suggests that the use of the word niggardly in this context most likely did reflect a stereotype about race, though not as might be expected, especially since the word has no linguistic meaning to race. These slips, along with illustrations generated in more controlled conditions, will be more fully analyzed below.

Closely examining negative unconscious linguistic references to race is im- portant for a number of reasons, not least of which is establishing valid – instead of ‘intuitive’ or seemingly commonsense – explanations. Establishing a slip as having no racial meaning is equally as important as establishing that a given slip does exhibit a negative attribution to race. A lack of proportion, by many Whites, for denying that anything racial, is reflected in such slips (see Van Dijk, 1992; Van Dijk, 2002). Conversely, many African-Americans perceive racism in all slips. It is suggested here that this lack of proportion is partly due to a lack of informed methods for analyzing slips about race.6 The social implications and consequences of unconscious linguistic references to race are clearly understood by most and thus will not be addressed here. The above slips raise a host of important questions. First, are they cognitively possible?7 Second, do such slips reveal an unconscious racism? Third, or are they simply linguistic errors caused

Haskell: Unconscious linguistic referents to race 61

by mechanical mistakes in the speech production process as has been suggested for the talk show host and athletic director slips previously mentioned (indeed, just as slips-of-the-tongue have historically been explained by most of linguistics, e.g. Fromkin, 1973)?8 Finally, are ‘slips’ more numerous than might be suspected?9

The question of method Though it may seem obvious that unconscious schemata about race exist as evidenced by slips and other forms, it is nevertheless necessary to approach the phenomena as rigorously as possible. Accordingly, this paper will: (a) bring two methodologically informed – but divergent – approaches to bear for (b) establishing that unconscious schemata about race do in fact occur in various forms, and (c) apply these methods to explaining, analyzing, and validating schemata about race. Together, the methods will show the interrelationship between the social and individual manifestation of racial ‘slips.’

Findings from these two methodologically informed approaches will be applied to apparent unconscious slips about race. The first is based on experi- mental designs (e.g. Bargh and Gollwitzer, 1994; Bargh et al., 1996; Duncan, 1976; Vanman et al., 1997; Wittenbrink et al., 1997). While experimental findings have been central in understanding stereotypic unconscious schemata about race and validating their existence, they have seldom been applied to every- day instances. This paper will apply the findings from this method to everyday schemata about race.

The second methodology applied here for recognizing, analyzing, and valid- ating instances from everyday settings is a novel psycho-linguistic and logico- mathematic controlling method. This method diverges from the standard experimental/statistical paradigm. A complete online explication of the method is available (Haskell, 2003a), along with a comparative analysis of the experi- mental/statistical paradigm (Haskell, 2003b).

Research has long suggested (Haskell, 1978, 1982, 1983, 1991, 1999, 2001) that linguistic referents in the content of verbatim protocols of verbal discourse in small social settings often map directly onto the narrative situation, yielding consciously unintended semantic referencing of group and interpersonal dynamics. I have termed these consciously unintended references sub-literal (SubLit), meaning – metaphorically – ‘beneath the literal.’ More recently, a theor- etical basis (Haskell, 2004) for the validation of such seemingly anomalous findings has been developed. Further, Haskell and Badalamenti (2003) found an algebraic structure under-girding a set of SubLit referents, further indicating their cognitive basis and veridicality.

Before analyzing the use of the slips ‘coon,’ ‘monkey,’ and ‘niggardly’ cited at the opening of this paper, it is first necessary to briefly outline the experimental methods and their findings, and to introduce an attribution taxonomy for distinguishing between schemata that are racial and those that are racist. The final section of this paper will present a series of illustrations from the psycho- linguistic and logico-mathematic methodology.

62 Discourse & Society 20(1)

Experimental methods and induced activation of schemata about race EXPERIMENTALLY ELICITING COGNITIVE SCHEMATA ABOUT RACE Unconscious stereotypic schemata about race has been experimentally established since at least the 1970s. A study conducted by Duncan (1976) showed students a video tape of a white man and a black man having a rather low-key argument. In one version of the video the white male was seen lightly shoving the black male with whom he was having the minor argument. In another version, the black male was seen lightly shoving the white male in the same scenario. In the video with the white male lightly shoving the black male, only 13 percent rated it a violent act; however, in the version where the black male was seen lightly shoving the white male, 73 percent rated it a violent act. These findings do not reflect outdated attitudes about race. Consider the two following studies.

Vanman et al. (1997) recruited white volunteers who believed they were not prejudiced in any way and connected them to electrodes that registered their minute facial muscle activity (under experimental circumstances such muscle responses are nearly impossible to consciously control). They were then shown pictures of both Caucasians and African-Americans and asked to imagine themselves interacting with their racial opposites and to select which ones they thought they might like the best. With most Caucasian subjects who selected African-Americans, their facial muscles belied their conscious beliefs – at least on some level. With these white subjects, when pictures of blacks were shown, their frown muscles registered more actively; this was not the case when shown pictures of white people.

This technique of measuring muscle activity is not new. Similar experiments have been conducted using gender and age as the focus and have shown the equivalent results. It seems unconscious intentions are telegraphed on many levels. Experiments like these demonstrate not only the existence of unconscious processes but also that these unconscious schemata can inf luence over t behavior and judgments. Many other experiments have shown that people with unconscious stereotypes and prejudice are more likely to activate their stereotypic schemata when shown ambiguous pictures, especially if they are angered or in a hurry, or otherwise feeling stressed. It seems when angered, in a hurry, or stressed, that people cognitively regress to their more unconscious negative schemata.

At this point, two crucial questions remain. First, where do such schemata originate? Do they originate within the individual or from a social/cultural matrix? Second, what do unconscious expressions of negative schemata about race mean? Do they mean the person is racist in an intentional and/or conscious sense of the term, or perhaps ‘only’ in an unconscious sense; or do they have meaning on yet other levels? The experiments described not only provide additional experimental evidence that schemata about race exist, but more importantly, address the question of what the expressed schemata mean.

Haskell: Unconscious linguistic referents to race 63

CONTROLLED VERSUS AUTOMATIC ACTIVATION OF SCHEMATA ABOUT RACE As noted at the opening of this paper, recent years have seen considerable development of experimental techniques for accessing unconscious attitudes and beliefs. One of these methods involves priming (Draine and Greenwald, 1998). Masked priming involves stimuli presented either outside of one’s focal attention, or presented so rapidly that the stimuli cannot be consciously perceived, but which, nevertheless, influence subsequent responses. Priming methods have stimulated an area of research on automatic versus controlled activation of unconscious stereotypic schemata.10

Automatic activation of negative stereotypes (in this case about race) are considered the consequence of being socially and culturally conditioned. Thus, automatic activation of negative schemata about race can suggest that a person either (a) did not consciously hold racist/racial beliefs, or (b) that the person held racist/racial beliefs but only on an unconscious level. Automatic activation of negative schemata about race with some level of controlled cognitive processing has been viewed as involving some level of individual intent, thus suggesting a personally held belief on some level. It should be noted that controlled activation does not necessarily mean conscious. Controlled activation is indicated by the degree of time delay in the activation measured in milliseconds. The longer the time delay, the more it is assumed that the activation is being unconsciously controlled. This research indicates that stimuli presented either during an experiment – or present in everyday environments – can elicit automatic and controlled unconscious stereotypic schemata.

On the basis of a psycho-linguistic and logico-mathematic methodology, Haskell (1986–7) reached, early on, a similar distinction between controlled and automatic activation of schemata of race, noting that, ‘Listening to apparent casual verbal repor ts in a systematic and linguistically informed manner continues to reveal both personally current and psycho socially vestigial remains of racial conditioning’ (p. 75). In automatic versus controlled terms, personally equates (roughly) to controlled activation; psycho socially vestigial indicates the remains of sociocultural racial conditioning of stereotypes that are automatically activated. It is always a positive sign in science when there are different methods and findings that corroborate each other.

Priming techniques have not only been utilized to activate stereotypic sche- mata on gender, social affiliation, and achievement behaviors (e.g. Bargh and Gollwitzer, 1994), but have also been widely used to automatically activate negative stereotypes about race (e.g. Devine, 1989). In an experiment by Wittenbrink et al. (1997), a group of volunteer subjects were primed by flashing either words or pictures onto a screen at a rate too brief for them to consciously perceive. Words or pictures were flashed that represented racial minorities. With a second group of subjects, no images were flashed during the presentation. The researchers then angered the subjects. They then asked them a series of neutral questions. Subjects who were primed with images of African-American males were more likely to respond in a hostile manner to the neutral questions than were those who were also angered but who were not primed with African- American male images (see also Bargh et al., 1996).

64 Discourse & Society 20(1)

As briefly mentioned above, this research suggest two kinds of stereotypic schemata, one cultural, the other individual or personal. It is generally accepted that automatically activated stereotype schemata are created and conditioned by cultural transmission, including mass media, etc. Accordingly, automatically activated cultural schemata are generally thought to imply no volitional and individual motivation involved in their activation. In contrast, controlled activation of stereotype schemata are viewed as the consequence of individually held beliefs and implies some level of individual volition and motivation involved in the activation of schemata about race (though not necessarily conscious). Such studies demonstrate human deception processes (see Smith, 2004).

Extending this research, Bargh and Barndollar (1996) suggest the automatic activation of what they term ‘chronic goals and motives’ for explaining unconscious responses to stimuli. Briefly, they view the unconscious as the repository of goals and motives that, with frequent use or repetition, become automatically activated strategies for responding to the environment. They suggest that under certain conditions, unconscious schemata subserving goals and motives can be automatically activated. This automation is analogous to the functioning of highly practiced motor skills (just as a ‘punch-drunk’ professional fighter may automatically begin to throw a punch at the sound of any bell). In a similar manner, environmental stimuli (e.g. the presence of a black person) can automatically activate the chronic goal or motive schema of white people. Bargh (1997) suggests that much of conscious everyday behavior is the consequence of these conditioned and automatically activated chronic goals and motives, activated by environmental situations functioning as priming stimuli. The effect of priming will be described when illustrations from the linguistic and logico- mathematic methodology developed by Haskell (2003a) are presented later.

It should be noted that an important unresolved issue in the literature is whether automaticity of unconscious motives and goals reflects a ‘fully’ automatically activated cultural stereotype or whether it also represents some level of individual or controlled processes (see Kunde et al., 2003). Though this problem has not yet been resolved, Bargh and Barndollar (1996) suggest the automatic versus controlled dichotomy may be a false one, with the activation of a stereotype having both automatic and controlled or motivated components. It may likely turn out, however, that some schemata are the consequence of automatic activation, some of controlled activation, and some a combination of both automatic and controlled. Before extending the analysis of the ‘slips’ cited at the opening of this paper, it is necessary to more closely define some terms.

An attribution taxonomy for ‘slips’ about race The terms ‘race,’ ‘racial,’ and ‘racist’ are typically used in widely differing ways. The term ‘race’ is used here as it is commonly understood by the majority of the population: as a genetic pool that distinguishes a group of people belonging to it from other gene pools. The term ‘race’ is used here only because it is the accepted medium of semantic exchange and therefore convenient in communicating the subject of this paper. The term ‘ethnicity’ is often used as a more socially

Haskell: Unconscious linguistic referents to race 65

acceptable equivalent for the term race. ‘Ethnicity,’ however, is a term denoting social, linguistic, and national affiliation, not only biological. Other than for com- municative convenience, however, the author considers the biological concept of ‘race’ to be a social construction and a ‘myth’ as the anthropologist Ashley Montague (1964) argued.

AN INITIAL ATTRIBUTION TAXONOMY The term ‘racial’ is developed here in four senses. The following distinctions evolve out of Haskell’s (1986) sub-literal methodological findings, as well as the automatic activation research, and are important conceptual distinctions both in terms of prejudice and in terms of cognitive structure. After this brief description of the three distinctions, they will be explained more specifically, as the illustrations presented in this article are methodologically analyzed. Let it be clear, however, that what is intended here when claiming that an apparent racist slur may in fact not be racist: the claim is not meant to explain away racial prejudice or to provide a rationalization for prejudiced beliefs, attitudes, and other racist slurs.

Racial1 refers to schemata about race that merely recognize ‘difference,’ in this case skin color. For example, Haskell (1986–7) observed that linguistic recognition of racial difference occurs in small social gatherings where the presence of only one or two African-Americans serves as a priming stimulus to Whites. This recognition of difference is often manifested by the appearance of various literal topics about the color black, e.g. chocolate ice cream, black eyes, and electrical black-outs. In such instances, there is no negative stereotypic schemata about race, only a recognition of difference. As a further example, in a news story describing Sullivan Phillips, a black man who, at the time, was one of only two black players for the Boston Bulldogs, it was said that the player’s ‘ability to change the complexion of the game was evidenced every time he stepped over the chalk line.’11

Literally, the description refers to the player’s ability to change the quality of the game every time he scores. Unconsciously, however, it likely indicates an unconscious recognition of race. According to most dictionaries, the first meaning of the word complexion is about skin color and texture, and a chalk line is, of course, white. So, it is reasonable to conclude that the description alludes to racial difference. There were many other non-color-associated words that could have been selected to describe this player’s impact on the game by his ability to score points.

Racial2 refers to automatically activated negative stereotypic schemata about race that are neither consciously nor unconsciously intended to devalue an individual or group, and which would not be consciously believed by the speaker. These would reflect a fully automatic activation of culturally conditioned stereotypes. For example, in situations where the presence of African-Americans may function as a priming stimulus, literal topics about biological differences (for example incompatible blood types, and different species not being able to mate) reflect incorrect beliefs and stereotypes about race (see Narrative # 6).

Racial3 refers to stereotypic schemata about race that are neither consciously nor unconsciously intended to devalue an individual or group but which may be consciously believed by a speaker as being ‘simply a social fact.’ For example,

66 Discourse & Society 20(1)

topics about animal pedigrees in reference to a light skinned black person, with an ensuing discussion of pure strains versus half breeds, or about the tobacco Half and Half. (Also see illustration # 1.)

Racial/Racist4 is used here to describe negative schemata about race that originate from some cognitive space with a personal intention (conscious or unconscious) of devaluing an individual or group and is thus racist; for example, literal topics about low IQ, hanging of things (lynching – see Narrative # 6), and other topics that involve extreme devaluing and violence.

People who hold Racial1, Racial2, and Racial3 schemata are, arguably, not racists (as the term is used here, and perhaps commonly understood by most Caucasians). Such automatically activated schemata may merely denote a person’s lack of close experience of being among people of a different race, and/ or lack of knowledge about race. In sum, schemata activated about race may be racial but not racist.12 As seen from the above scheme involving the four senses of the term racial (and further below), to assess whether or not a literal story is an unconscious racist reference requires considerable social context, history, and individual information regarding the person who generated the topic.13 It is suggested along with the above research that this brief attribution taxonomy will provide a useful frame within which to analyze activated schemata about race.

Analysis of the opening ‘slips’ about race in the news media This section analyzes the opening illustrations of unconscious ‘slips’ about race informed by the two methodological frameworks outlined above. Was the radio talk show host’s intended remark (that Secretary of State Condoleezza Rice’s possibly becoming football commissioner would be a ‘big coup’ or boon to the sport, but which resulted in a ‘slip’ saying that she would be a ‘big coon’) racial or racist? Similarly, are the two ‘slips’ by the high school athletic director and the weather forecasters (referring to ‘Martin Luther Coon Day’) personal unconsciously slipped schemata about race, or just speech errors? Further, did they reflect an automatic or controlled activation of an unconscious schema about race?

First, according to most linguistic research on speech errors, the above illustrations would be considered the consequence of mechanical errors in the speech production process (for example, it is known that low-frequency words are ‘slipped’ more often than are high-frequency words). Indeed, it was suggested that the talk show host’s slip was simply the result of ‘morphing’ the words ‘coup’ into ‘coon.’ Linguistically, the talk show host’s slip could theoretically fit the speech error model more than the athletic director and weather forecaster’s slips, as thinking that an African-American (and female) becoming football commissioner would be a big ‘coup,’ to the sport is phonetically closer to ‘coon,’ than is converting ‘king’ to ‘coon.’ It might be noted, too, that if the coup-to-coon slip was verbalized in a non-race setting where racoons were contextually active, then it would not be reasonable to assign the ‘slip’ a race-based meaning, but rather to consider it simply a speech error. Unless otherwise justified, this paper will not consider such ‘slips’ as merely speech errors.

Haskell: Unconscious linguistic referents to race 67

Without further knowledge to the contrary about the talk show host and the athletic director’s behavioral and attitudinal history regarding race issues, there is no way to adequately assess here whether the slips are automatic or controlled activations reflecting Racial/Racist4 (being the result of a controlled activation of a negative schema about race). At the least, however, they likely reflect automatic activations reflecting both Racial2 and Racial3 categories. One can reasonably assume that if Secretary of State Rice were white, this linguistic ‘slip/error’ would likely not have occurred. If, however, personal knowledge about the talk show host, the athletic director, and the weather forecaster showed a personal history of devaluing minorities, then with this context it would be reasonable to conclude that the ‘slips’ revealed racism on some level (automatic and/or controlled activation).

In terms of contextual background, there is some per tinent personal information about the weather forecaster. At least one African-American who knew him reportedly said, ‘As a black person, I can say he is anything but racist. This guy has a history of working for and helping minorities.’ With this context, the weather forecaster’s slip is likely a Racial2 category slip, though it still could be a Racial3 but perhaps not Racial/Racist4.

One could argue that the adjective ‘big’ (in describing Rice’s becoming commissioner being a ‘big coon’) is linguistically significant in that the term ‘coon’ historically and typically referred to large black males. Why not have selected the word real in relation to ‘coup’ and ‘coon’ as the ‘slip?’ Linguistically, the literal consciously intended meaning of a ‘real coup’ would have connoted that same meaning. The field of linguistics accepts that most lexical selection is not only an unconscious process, but that it is rule-driven if not ‘lawful.’ Consequently, it is reasonable to assume that selecting the adjective ‘big’ in the context in which it was used was not random. It is tempting to hypothesize that this schemata about race reflecting the extreme derogatory meaning it expresses as likely having at least an element of controlled processing and therefore assigning it to the Racial/ Racist4 category. But not having adequate evidential support here to do so, and given the seriousness of attributing racism to a person, it should be left to future findings about the speakers to resolve the issue.

In CNN anchor Lou Dobbs verbal ‘slip,’ ‘These cotto . . .’ (the full meaning of which would have been ‘These cotton picking politicians’ instead of what he ‘meant’ to say, ‘These ridiculous politicians . . .’), does the slip reflect an automatic or a controlled activated racial schema? Not having sufficient information about Dobbs’ history, it is not possible to assess the exact meaning of his slip. However, at the least such slips demonstrate a continuing presence of automatically activated unconscious social/cultural stereotypic racial schemata.14

In the Howard Cosell example, where he described the African-American Washington Redskins wide receiver, Alvin Garrett, running down the field by saying, ‘Look at that little monkey run,’ there is sufficient personal knowledge to suggest that this was neither a Racial1, Racial2, Racial3, nor a Racial/Racist4 reference or slip. Cosell apparently not only used this phrase frequently in his own family to describe children running, but to describe other white players. For example, there is apparently 1972 video footage of Cosell refer ring to

68 Discourse & Society 20(1)

Mike Adamle, a 5-foot-9-inch, 197-pound Caucasian, as a ‘little monkey.’15 What may appear to be a racial and/or racist reference, then, may in fact not be.

The immediate response to see Cosell’s remark as revealing unconscious racism demonstrates the need for informed analyses. Though arguable, the perception of racism attributed to this incident was in part likely responsible for Cosell’s no longer hosting Monday Night Football shortly after the incident. In the absence of such analyses, ‘rumors of racism’ may persist where it does not exist. For example, in an otherwise informative recent paper (Salgado, 2006), the author simply repeats the original racist attribution to Cosell’s remark.

In the 1999 mass media example about a racial incident involving the use of the word ‘niggardly’ in reference to a budget by a white aide in the office of Washington DC African-American mayor Anthony Williams, some black members of the meeting immediately walked out in protest upon hearing the word. In another context, the word would not be newsworthy, despite its seeming association to the so-called N-word (i.e., ‘Nigger’), even taking into consideration these ethnically sensitive times.

The word elicited considerable national media attention, including two pieces in the New York Times, one by a reporter, and an Op-Ed piece by Pinker (1999), a widely respected cognitive scientist cum linguist.16 Most of the press, along with Steven Pinker’s Op-Ed piece, all disavowed not only any racial (stereotype) but also any racist (negative) intent by the white aide. More importantly, the NAACP issued an almost instant press release agreeing that the white aide’s use of the word niggardly had no racial or racist meaning.17 In one news report, Chairman of the NAACP, Julian Bond, reportedly also disavowed the incident, saying ‘This whole episode speaks loudly to where we are on issues of race. Even “imagined slights are catapulted to the front burner.’’’18

In disavowing any references to race, the reasoning – correctly – noted that the word ‘niggardly’ means ‘miserly’ and has absolutely no grammatical linkage or linguistic history that would suggest anything racial. Indeed, using the word was simply coincidental – or so it might seem on a superficial analysis. Accordingly, the African-Americans who thought the term was a racial slur were considered overly ‘paranoid,’ or at the very least overly sensitive.

Most linguists agree a large part of everyday language including accessing and word selection is accomplished unconsciously. This is especially the case with complex grammar (out of the many hundreds of grammatical rules used everyday, most cannot recite many of them). As the highly respected psychologist cum linguist, Pinker, correctly pointed out, any word selected during a conversation occurs in a two stage process. First, because words have multiple meanings, the brain reviews (called lexical accessing) all possible meanings of a word from a person’s past experience. Then, from the context, the brain chooses (called lexical selection) out of all the possibilities the assumed appropriate meaning in the context of its use. In addition, Pinker correctly notes that the ‘ar’ in the word nigg(ar)dly is often pronounced ‘er’ as in nigg(er)dly. Pinker uses these linguistic facts and analysis, however, to deny that any racial meaning was present.

Haskell: Unconscious linguistic referents to race 69

From the perspective of this paper, the significant question not addressed about this incident is why the particular word ‘niggardly,’ with its similar morphology to ‘nigger,’ was selected in this particular context as opposed to its many other, and higher frequency, synonyms? Indeed, why was miserly not used to describe the budget, or stingy, or tight, or tightfisted, or penurious, or penny- pinching, or frugal, and so on? Indeed, this is the significant question that must be answered in order to understand not only the racial implication of the incident in Mayor Williams’ office, but all other ‘slips’ as well.

The term niggardly voiced by a white aide was likely at least an automatic activation of an unconscious reference to the fact that the Mayor and most of his advisors were African-American. In disavowing any racial meaning to the word niggardly, the analyses were technically correct as historically the origins of the word carries no racial meaning. But socio-culturally, the disavowal was likely not correct. As with most of the above examples, the word niggardly likely reflected at least a situationally primed automatic activation of a sociocultural schema of at least a Racial1 and perhaps a Racial2 or even a Racial3 category. That the white aide was selected to work with Mayor Williams could argue against the use of the term reflecting Racial/Racist4.

Sub-literal (SubLit) references to schemata about race Unlike ‘slips,’ contents of some verbal discussions that are clearly not about race on a literal level, sub-literally yield referents to race issues extant in the conversational setting (Haskell, 1986).19

Methodologically, there are four fundamental questions that are seldom asked but which must be answered about verbal narrative content in social settings:

1. Out of all the possible topics or stories in verbal discourse, why is a particular topic or story introduced into a conversation?

2. Why, out of all possible times or occasions or circumstances, is a topic or story introduced into a conversation at a particular point in time?

3. Why, out of the many stories selected into the conversation, is one often sustained and elaborated upon either by an individual or by others in a conversation?

4. Out of all possibilities, why is a particular wording, phrasing, order structure, or syntax used?

The short answer to these questions from a SubLit perspective is that the selected nar ratives allow the unconscious expression of af fective schemata about the social setting – in this case about race. It is to these illustrations of SubLit expressions of otherwise publicly unexpressed affective schemata about race that this paper now turns.20

SubLit narratives about race In part, because the idea of unconscious meaning is almost reflexively associated with Freud and thus with virtually omni-meaningful interpretations, the series

70 Discourse & Society 20(1)

of illustrations below will open with a fairly complex and systemic one where the methodological procedures involved in validating SubLit meanings will be described in some detail. This illustration will thus function as a kind of methodological exemplar or paradigm case for the less methodologically detailed illustrations.

Multiple illustrations from both laboratory and everyday settings are presented to provide a range of the kinds of topics, stories, and language that are used to express and construct SubLit references to race. It should be noted that cognitively there is no difference between illustrations from a laboratory setting and those from everyday situations. Methodologically, however, in analyzing narratives generated under laboratory conditions, the advantage is having the history and contexts of the sessions. In addition, the discussion can be tape- recorded and transcribed verbatim for a more controlled analysis and validation. For practical purposes, since narratives from the more context documented laboratory conditions have been repeatedly validated with the methodology, it is reasonable to infer that findings of many similar narratives from everyday settings can be assumed, at least initially, to have validity as well.

Sub-literal meanings are basically derived by (a) mapping and matching the narrative content onto the narrative situation, along with (b) understanding the context(s) of the setting.21 Other procedures are briefly cited that can be found in Haskell (2003a). Though there are other procedures, once the two basic procedures of mapping and matching are understood, the meaning of most SubLit references are often virtually transparent (as will be seen below). Finally, it should be emphasized that the following SubL references to race should be viewed within the automatic versus controlled and priming research frameworks described above, with conditions within the social setting functioning as priming stimuli.

NARRATIVE # 1

Narrative A discussion was initiated about /Dogs That Were Not Too Smart/because they had been /Deprived/ as pups. The speaker referred to a specific breed of dog: /A Black Labrador Retriever/ that the speaker once had who was /Retarded/. It was further stated, however, that it was a /Gorgeous/ dog.22

Context (a) The group was composed of all white members except for one older African- American woman. The nar rative about dogs was introduced by an older white woman. The black woman was perceived to be having more difficulty understanding and analyzing the group dynamics than the rest of the discussants. When this became publicly recognized, and after a brief flurry of disjointed repartee, the above narrative was initiated by the white woman.

SubLit meaning At this point one might hypothesize that the literal narrative is a SubLit reference to the black woman, especially after mapping the content and context of the narrative onto the group situation. Mapping the topic /Dogs That Were Not Too

Haskell: Unconscious linguistic referents to race 71

Smart/ because they had been /Deprived/ as pups, onto the narrative situation – along with the context of historical and extant negative stereotypes about African-Americans – suggests it was a SubLit reference to the black woman being perceived as experiencing difficulty learning, i.e. retarded. It is still a wide-spread negative stereotype among segments of the white population about African- Americans that their childhood is one of social deprivation, which, it is believed, causes blacks to have lower intelligence than Whites, and thus more difficulties with learning.23

Further validation procedures and operations In addition to mapping and contextual procedures, there are several significant methodological operations and linguistic linkages attached to this SubLit reference which suggest the narrative referenced the black woman. First, the association to the color black is, of course, congruent with the black woman being a priming stimulus. Second, the dog was said to be retarded. Third, the dog was a /Retriever/. That it was a /Retriever/ involved semantic linkages to racial stereotype schemata. The term has historical semantic associative linkages to servant/slave behavior. That the dog was specifically a /Labrador/ retriever is likely a reference to the course being known as a laboratory course, and thus a linguistic linkage between the literal topic and the narrative situation ([3.4.] Transitional Linkage Operations; [8.4.] Paronymic Operations).

Methodologically most interesting is that the white female described the dog as /Gorgeous/. This lexical item is important as it also indicates a linguistic paronymic linkage ([8.4.] Paronymic Operations) to an earlier session in which the black female had given the white female an ashtray she had made for her. The white female accepted it, describing it as /Gorgeous/ (and thanked the black woman as one would thank a small child). It had also been stated by the black female that she was originally from /Georgia/. These linguistic linkages are considered the SubLit equivalent of paronym relations. Paronymic words derive from the same root or stem as another word, like meanie and meaning. Words that are graphically perceived to be paronymic-like are used to create SubLit referents.24 Recall the ‘nigg(ar)dly’ illustration. See Figure 1 overleaf.

Racial or racist analysis Having methodologically established that this literal narrative is a SubLit sche- mata about the black woman in the discussion, the question to be answered is whether or not the white woman’s narrative reflected a racial or a racist schemata? The context and history of this narrative shows that the white woman was a self-professed ‘liberal’ and was quite active in civil rights organizations. This would suggest that she was not consciously prejudiced in the sense of being Racial/Racist4 though clearly she possessed stereotypic sociocultural schemata about race in the sense of Racial2 and perhaps Racial3.

Thus the narrative likely reflected an automatic activation of culturally conditioned ‘racial’ schemata. But more than this, it is also likely that, though she was not racist in the typical sense, she may have consciously believed in some of the automatically activated stereotypes, i.e. early social deprivation resulted

72 Discourse & Society 20(1)

F I G U R E 1 . Black labrador retriever matrix.

Haskell: Unconscious linguistic referents to race 73

in many blacks being slow learners. This is perhaps contextually evidenced by (a) her thanking the black woman for the ashtray as one would a small child, (b) because of the coherent complexity of the SubLit racial schemata, and (c) the personal nature of it.

It is assumed here that the complexity of the narrative schemata involves some level of controlled cognitive processing. Thus, this constitutes belonging to the Racial3 category but not Racial/Racist4. Finally, linking SubLit narratives to the laboratory experiments, the fact that the black woman was the only African American in the discussion was a physical priming stimulus that automatically activated the negative racial stereotypic schemata.

NARRATIVE # 2

Narrative In a transition of one network television business news show to another, the commentators repeatedly used the phrase /Holy Mackerel, Andy/ as well as the word ‘Buck’ in various contexts.25

Context (a) The show commentators were all white males. Seldom, if ever, had there been a black professional man associated with the show. The author suspected, and indeed predicted, that the references to/Holy Mackerel, Andy/ and ‘Buck’ were SubLit references to an African-American, but there was no black news commentator or guest on the current show. From a SubLit perspective, this apparent prediction failure was disappointing. In continuing to watch the business news show, however, the prediction was validated.

SubLit prediction An advertisement for the program to follow came when the anchor introduced one of his professional guests: a rather large and quite dark-skinned black man with stereotypical African-American features (broad nose and large lips).26 The SubLit analysis correctly predicted the appearance of a black man. Clearly, the commentators on the previous show were aware of the black man waiting in the studio and recognized the rare presence of a black fellow, especially as a professional guest.

Context (b) For readers who may not be aware of the history around the phrase ‘Holy Mackerel, Andy’ and word ‘Buck,’ and who are wondering what relevance the phrases have in hypothesizing that they were references to race: first, the term ‘buck’ was historically, and continues to be, used pejoratively to describe large black men.27 Second, the phrase /Holy Mackerel Andy/ was a well-known phrase used in the now classic old radio show ‘Amos and Andy,’ where two white men presented themselves via a black patois as black men. During the show, in response to some story or conversation, Amos would repeatedly say to his partner, ‘Holy mackerel, Andy!’ In fact the phrase was a kind of advertising trademark

74 Discourse & Society 20(1)

of the show. (Parenthetically note that this is an example for the importance of having a large knowledge base about race for recognizing SubLit references.) It is significant, too, that the way the anchor enunciated the phrase was exactly the manner in which it was done on the Amos and Andy show.

Racial or racist analysis Clearly, this narrative reflects an automatic activation of a sociocultural schema about race. Again, without more personal information about the program anchor, it is not possible to make a methodologically informed judgment about whether this narrative reflects a Racial/Racist4 schema; however, it likely reflects at least a Racial2 and perhaps a Racial3 category schema.

NARRATIVE # 3 Watching a documentary on the 1974 Mohammad Ali–George Foreman fight in Zaire, Africa, Norman Mailer, the US writer, novelist, and essayist was extensively interviewed for his commentary on the fight as he had been in Zaire for the championship match.28 During Mailer being interviewed, SubLit references to race were evident (though no one, of course, recognized them in the documentary).

Narrative After showing and describing Foreman repetitively hitting a punch bag (the big ‘heavy bag’) during practice in the same spot with great force, Mailer said that Foreman’s great punching strength left the bag with an indentation the size of a . . . . . . . . . (fill in the blank).

Context Any number of words were possible to describe the size of the indentation. Since it was a sports event, Mailer could have compared the indentation to the size of a softball, or perhaps soccer ball, or a large cantaloupe. But he didn’t. The fight was, of course, between two African-American men.

SubLit meaning Mailer described the indentation as ‘Half the size of a small watermelon.’ This a SubLit reference to race. Recall that historically, eating watermelon is associated with African-Americans in the US. So, it must asked again: of all possible objects, (a) why was a fruit selected to compare the size of the indentation, and (b) why the particular fruit watermelon? Why not a cantaloupe?

SubLit Operations Methodologically more important than simple context for assessing the validity of Mailer’s SubLit meaning, note that Mailer had to make the watermelon stereotype fit the size of the indentation. The indentation was not in fact as big as a typical watermelon. But nor was it even the size of a small watermelon.29 It was, he said, ‘Half the size of a small watermelon.’ It is a very peculiar phrasing to say that the indentation was not only ‘half ’ the size of a watermelon, but it was ‘half ’ the size

Haskell: Unconscious linguistic referents to race 75

of a ‘small’ watermelon. What is significant here, is that cognitively Mailer had to linguistically gerrymander the size of the watermelon to make it fit the actual size of the indentation. Why not have said, twice the size of a grapefruit? Finally, why not have used a sports-related image like a softball, volleyball, etc.?

Racial or racist analysis Does this narrative by Mailer reflect just an automatically activated racial schema or does it reflect a more controlled racist schema on some level? Knowing that over the years Mailer has been known for being ‘liberal’ and ‘hip,’ and not for his racist views, it would suggest his statements were racial but not racist. However, methodologically it is important to point out that the linguistic gerrymandering involved in the phrase ‘Half the size of a small watermelon’ would suggest some level of controlled processing. As in NARRATIVE #1 earlier, it is likely that, although he does not hold racist views in the typical sense, he may unconsciously believe in some of the automatically activated stereotypes, and thus it likely reflects a Racial3 category.30

NARRATIVE # 4

Narrative After a little chitchat with a colleague and his wife, without apparent continuity with the previous topics, a man said that his ‘cousin had recently adopted a child from Africa.’ He emphasized that his ‘cousin’s decision was unwise because he’s really too old for that sort of thing.’

Context The colleague is white and middle-aged and his wife is a black Jamaican 20 years younger than he. It is still the case for many people that a middle-aged white man married to a young black woman is highly charged and activates a variety of stereotypes, fantasies, and suspicions, although seldom mentioned in polite conversation.

SubLit Meaning This story reflects a concern with both age and race: first, it is significant that the statement by the speaker that his ‘cousin had recently adopted a child from Africa,’ references Africa. Second, that it references a young child. Thus the statement is SubLit for his feeling that the white colleague’s Jamaican wife was young enough to be his adopted daughter; the reference to Africa is transparent. The speaker’s statement that his ‘cousin’s decision was unwise because he’s really too old for that sort of thing’ is SubLit for his feeling that the colleague was too old to have such a young wife.

Commentary Again, it might be asked why this particular story was selected into the conver- sation? The answer is the story reveals that for some people both a difference in age and in race remain significant issues.31

76 Discourse & Society 20(1)

Racial or racist analysis Given the above, did the narrative reflect just an automatically activated racial schema or does it reflect a more controlled racist schema on some level? Not knowing more about the personal history of the speaker, it is not possible to ascertain. It certainly reflects a disapproval of the age differential. It would seem from this narrative – because race was included by the adoption having taken place in Africa – that it likely reflected Racial2 and Racial3 categories.

NARRATIVE #5

Narrative Another man talking to the author’s colleague and the colleague’s wife (as above) talked about Louis Jordan’s music and specifically mentioned a song entitled ‘That Chick’s too Young to Fry.’

Context Same context as above. In addition, Louis Jordan is an African-American blues singer, and the lyrics to the song mentioned are considered obscene by many.

SubLit Meaning Clearly, mentioning the song ‘That Chick’s too Young to Fry’ not only reflects a SubLit concern with the young age of the colleague’s wife, but with sexuality as well.

Racial or racist analysis Once more, SubLit analyses show deep schemata to race, and to wide age differen- tials between male and female. Did this narrative reflect just automatically activ- ated racial schema and sexuality schemata, or does it reflect a more controlled racist schema on some level? Once more, not having more contextual information about the personal history of the speaker, it is not possible to ascertain. Like the illustration above, it clearly reflects a disapproval of the age differential. As with the above example, too, it would seem that this narrative – because race was integrally included by the selection of a black composer – likely reflected at least a Racial2 and Racial3 category.

NARRATIVE #6

Narrative While standing in front of a man in line at a hardware store who was purchasing some big hooks with which to perhaps hang large objects in a shed or garage, the man in front of him said, ‘That looks like a man who’s going to do some hanging.’

Context The man purchasing the hooks was a very well dressed black man in a business suit and tie. The man making the comment was a working-class white fellow dressed in scruffy work clothes. The white fellow glanced behind him at the very well dressed black man. A fleeting look of surprise on his face was evident at

Haskell: Unconscious linguistic referents to race 77

seeing this well dressed, coiffed, and manicured middle-class African-American male. It was one of those awkward social encounters when two strangers make eye contact that calls for some kind of social recognition. Having made the eye contact, the white man clearly felt socially compelled to say something, some little ritual verbal formula to acknowledge the other’s existence, just as we ritually say ‘good morning’ to a person when first meeting them. On a literal level, the statement meant that it looked like the black fellow was going to be hanging some things on a wall somewhere.

SubLit Meaning Sub-literally translated, the statement, ‘That looks like a man who’s going to do some hanging,’ was at the least an association to the historical lynching of black males. More likely perhaps, it sub-literally meant: A black man who looks and behaves superior to a white man may find himself being lynched or hanging from a tree. (Ironically, this illustration is perhaps the one which receives the most incredulous responses. Yet, it is the one that almost invariably people spontaneously recognize as the SubLit meaning – before disavowing their spontaneous recognition of its meaning.)

Commentary Again, as with all SubLit conversations, it is methodologically important to ask why, out of a practically infinite number of responses the white man may have made, was the phrase ‘That looks like a man who’s going to do some hanging’ selected? Moreover, why were the ‘hooks’ selectively perceived out of a number of articles on the counter? Further, out of a nearly infinite number of syntactic arrangements of words in which ‘hanging’ could have been used, why was this particular arrangement of words selected? Why not, for example, ‘What are you going to hang with all those hooks,’ or ‘Those hooks work well,’ or ‘That looks like a man who’s going to hang some large implements,’ or some other similar phrase? The answer is that any of these alternatives would not have corresponded to the white man’s affective schemata about race. The phrase ‘That looks like a man who’s going to do some hanging,’ is an example of an ambiguous sentence whose surface structure has more than one underlying cognitive representation.

Racial or racist analysis Once again, whether this was a racial or racist schemata is not possible to say with any kind of methodological certainty. Strictly speaking, not having more context and personal history on the white man, there are no valid grounds on which to make such an attribution of racism. However, given the particular wording that was used, it is reasonable to assume that it was at least an automatic activation of a historical Racial3 stereotypic schema, and perhaps a Racial/Racist4 one due to the violent SubLit imagery involved.

Indeed, it is tempting to hypothesize that schemata about race that reflect the degree of hostility and violent imagery as this one about ‘hanging’ has at least an element of controlled processing. Accordingly, it should therefore likely be

78 Discourse & Society 20(1)

assigned to the racist category. However, not having other evidential support, and given the seriousness of attributing racism to a person on this basis, it is better left to future similar findings to resolve.32

Conclusion and implications Establishing that ‘slips’ about race have unconscious meanings benefits from applying multiple methods. This paper has brought two methodological areas of research to bare on understanding what are widely perceived to be unconscious racial ‘slips’ in everyday social settings. The findings here have shown that sometimes linguistic referents about race reflect no negative sociocultural stereotypic schemata; that sometimes they reflect an automatic activation of either culturally conditioned stereotypes, or individually motivated schemata, and at other times schemata reflect a conscious or unconscious racism.

Bringing two methodological frameworks together for explaining and analyzing schemata about race provides a systematic two-pronged approach to the phenomena, with each approach having its methodological strengths and weaknesses.33 The priming technique for inducing the automatic activation of negative attributions to race has proven extremely productive for experimentally establishing the existence of unconscious schemata about race, but it has been virtually limited to laboratory settings. In addition, assessing the ‘meaning’ of the findings tends to be problematic. Typically, questionnaires and racism scales are used for providing a kind of base line and context for ‘interpreting’ the meaning of the activated schemata about race.

In an age where overtly expressing negative attitudes about race is socially unacceptable, findings from intrusive questionnaires and racism scales are problematic. The advantage of the SubLit approach presented here is that it is an unobtrusive method (see Crosby et al., 1980) for accessing and recognizing negative schemata about race. Questionnaires require a level of conscious awareness of what they are to measure, thus rendering the validity of the findings questionable. Also, they are not sufficiently sensitive for distinguishing automatic from controlled processing, but more importantly for evaluating whether the findings reflect racial or racist meaning. Scores from racism scales may also contradict subject’s activated schemata and/or results from other more valid methods, for example, EMG facial frown muscles. Ingenious experimental designs continue to be developed to resolve some of these issues.

Finally, the SubLit approach turns everyday social settings into a natural laboratory, and thus has what Neisser (1976) terms ‘ecological validity’ i.e., everyday relevance. Unlike experimental designs, a sub-literal method is also able to analyze the complex psycho-linguistics of narratives.

In closing, perhaps the findings in this article will contribute to: (a) recog- nizing the continued force of both individual and social unconscious schemata about race; (b) providing a methodological approach; and thus (c) a more methodologically informed response to ‘slips’ about race; and (d) a more socially productive understanding of race issues on personal, cultural, and social levels.34

Haskell: Unconscious linguistic referents to race 79

A C K N O W L E D G M E N T S

I wish to express my appreciation to Dr Aaron Gresson and Dr David Smith for reviewing a draft of this manuscript.

N O T E S

1. Available at: http://news.bbc.co.uk/2/hi/americas/4842238.stm Also http://www. foxnews.com/story/0,2933,188830,00.html

2. Dallas County News (2005). Available at: http://www.dallasnews.com. Also see: h t t p : / / w w w. r e v i e w j o u r n a l . c o m / l v r j _ h o m e / 2 0 0 5 / Ja n - 1 8 - Tu e - 2 0 0 5 / n e w s / 25683849.html

3. http://www.reviewjournal.com/lvrj_home/2005/Jan-18-Tue-2005/news/2568 3849.html on KTNV-TV, Channel 13, weather forecaster Rob Blair. For video clip see: http://youtube.com/watch?v=GgEBHqeEbcI

4. Wolf Blitzer, CNN (2008). Situation Room, Friday 28 March, about 6:03pm. Video clip: http://www.youtube.com/watch?v=nofy3262nEA. At the same time, the immediate and widespread media context about the African-American, Barak Obama, running for US president was prominent. The political success of this black man is racially complex with some attributing his success to him not reflecting the sociocultural stereotype of a black man, i.e. that he is in effect a ‘white man’ with dark skin. His success will likely engender a ‘re-write’ of the psychosocial meaning of race and prejudice in the US, ironically on both sides.

5. Washington Post (24 April 1995). ‘Famous Howard Cosell Quotes,’ available at: http://www.washingtonpost.com/wp-sr v/spor ts/longterm/memories/1995/95 pass13.htm

6. The author is aware that many Whites as well as African-Americans would rather not call attention to such racial slips. As a society, however, we can not afford to ignore their effects.

7. Though decreasing, few areas in psychology have been more controversial than the veridicality of unconscious semantic meaning. The controversy, however, is no longer about whether unconscious cognitive processes exist, but about how ‘psychodynamic’ or analytically sophisticated they are. The voluminous literature cannot be detailed here (but see Dijksterhuis and Nordgren, 2006; Dijksterhuis and Van Olden, 2006; Haskell, 2002; Hassin et al., 2005; Lewicki, 1986; Merikle, 1998; Reber, 1993; Smith, 2004; Velmans, 1991 for research that can be seen as compatible with this paper, though the methodological findings presented here stand on their own).

8. With rare exceptions, there is little suppor t for an unconscious meaning in slips-of-the-tongue in cognitive science. However, Baars et al. (1992) and others (Sellen and Norman, 1992; Stemberger, 1992) suggest that current laboratory conditions may not be adequate for replicating the psychological processes of slips that occur in everyday situations. Baars et al. (1992: 308) conclude that ‘From some of our findings . . . the most immediate conclusion might be that the Freudian hypothesis is just plain wrong. But that is too simple.’ Based on their findings, Baars leaves the door open for methods that may be less ‘blunt’ (as he put it) than current ones. It should be noted that because many slips (likely even most) can be explained as mechanical or linguistic er rors in executing the lexical, phonological, and syntactic systems, it does not logically follow – as the research literature suggests – that all slips are simply speech errors.

80 Discourse & Society 20(1)

9. While the author is not aware of a database on the prevalence of ‘slips,’ with now recognizing ‘slips’ via extended seemingly literal narratives (see Haskell, 1986–7 and next section), whatever the prevalence may be thought to have been, it is clearly increased.

10. For a short cut to the automatic versus controlled activation literature, see the work of John Bargh. Available online: http://pantheon.yale.edu/~jab257/labhome.htm Also, the work of Patricia Devine. Available online: http://psych.wisc.edu/devine/ bio/pubs.htm

11. Bulldog vs. Barons: Bowditch Stadium, 7:30pm, PHILLIPS LEADS CHARGE, Bulldogs Look for berth in U.S. Open Cup by Damian Vega. MetroWest Daily News. Available at: http://www.townonline.com. I would like to thank my former student, Jennifer Reil, who gave the article to me, 25 September 2001.

12. Depending on how automatic versus controlled, i.e. social versus individual, activation is resolved regarding their being activated simultaneously in any given case, the distinction between racial and racist could be seen by some as distinguishing between levels of racism.

13. Context is a complex and often ambiguous concept. For an extensive analysis of the concept, see Van Dijk (1997, 2006). Also available online: http://www.discourses. org/download/articles/ For a logic-in-use meaning of context as applied to this paper, see Haskell (2003a): [2.] Contextual Procedures.

14. Apparently, CNN deleted Dobbs’ ‘slip’ in their printed transcript. See: http://gawker. com/374166/lou-dobbs-almost-calls-condi-rice-cotton+pickin

15. Shapiro, L. (1995) Washington Post Staff Writer. See: http://www.answers.com/ cosell&r=67

16. Again, see: http://pinker.wjh.harvard.edu/articles/media/1999_02_02_newyork times.html

17. See http://www.adversity.net/special/niggardly.htm 18. New York Times, Feburary 7, 1999. See also: http://query.nytimes.com/gst/fullpage.

html?res=9E06EFDC1738F932A05752C0A96F958260 19. All illustrations in this article reflect African-American and White relations. This is

because the populations that I have lived and worked in have not been diverse enough to provide SubLit narratives about other minority groups.

20. Other illustrations can be found in Haskell (1986–7, 1999, 2001) in various forms and levels of analysis.

21. The following procedures are a part of all SubLit analyses, and provide the basic access analyzing narratives; they will not be specifically mentioned further. They can be found in Haskell (2003a): [1.1.] Matching Operations. [1.2.] Isomorphic Mapping Operations. [1.3.] Alignment. [2.] Contextual Procedures. [2.1.] Current and Historical Contextual Data. [2.1.1.] Cultural Contexts. (see [2.3.] Expectations). [2.1.2.] Narrative Situation Contexts. [2.3.] Expectations. [2.4.] Knowledge Base.

22. This illustration from a small group dynamics training laboratory session was presented in a much abbreviated methodological form in Haskell (1986–7).

23. Notated in analogical form, it is as follows: Early deprivation: the black Labrador retriever’s later learning inability: the early

deprivation of the Black female’s life: her perceived difficulty to learning in the group. 24. Interestingly, whenever I have presented these SubLit references about race to an

audience, almost invariably most Whites are incredulous, whereas most African- Americans are not.

25. On FOX.7/10/04 (Bulls and Bears) Program leading to the following Cavuto Business Show.

Haskell: Unconscious linguistic referents to race 81

26. The black man is founder and CEO of a Wall Street firm who now appears regularly on the FOX program Cost of Freedom Business Block. For photo see: http://www. foxnews.com/story/0,2933,194175,00.html

27. Pete Williams (27 July 1999). NBC Nightly News, 6:45 pm, EST. Contrary to some reviewer comments, the stereotype to ‘bucks’ is not an antiquated one. In reporting on locating the abortion clinic bomber, Eric Rudolf, Pete Williams interviewed Mr. Rudolf ’s sister-in-law, Deborah Rudolf. She said that whenever he saw an interracial couple, where the female was white, he would remark, ‘Look at that big black buck.’ For more on the current use of the term ‘buck’ in relation to black men, see Haskell (2001) Chapter 9.

28. NBC, 9:30 PM EST, Sept 4, 1999. 29. Peruse old books about African-Americans and you will find that blacks and water-

melons are closely associated. Interestingly, and ironically, watermelons were indigenous to Kalahari Desert. They were brought to North America by European colonists and African slaves. The stereotype of watermelons is not dead. A search on the internet will yield numerous references to this stereotype.

30. Mailer was born in 1923, growing up in an era of blatant overt racism and therefore cognitively incorporated, on some level, racist and racial stereotypes about African- Americans.

31. I wish to thank my friend and colleague David Smith and his wife Subrena for these two personal examples.

32. I should note that most of the schemata to race that I have collected do not reflect the degree of violent imagery as this one about lynching. This may be artefactual to my not having been around people who are that deeply racist.

33. Van Dijk (e.g. 1997) conceptualiz es what this paper calls cognitive schemata as mental models. Both are a kind of internal map of situations, events, etc., of perceived external reality, and involve an organization for quickly processing experiences; both typically involve episodic (personal) memory. Mental models, then, like schemata, may either be about individual/personal realities or perceived sociocultural realities, e.g. in terms of this paper, racial stereotypes and ideologies extant in the culture or beliefs held by an individual. Sociocultural models may be transformed into personal/individual mental models. Depending on the particular conceptualization of Mental models, they may be either conscious or unconscious. Either way, they tend to be ‘ready made’ representations for dealing with situations, people, or events, and thus often preclude careful reasoning. In this sense they are ‘automatic’ activations of mental modules. Finally, context influences what mental models are activated and how they serve to process information. As with the racial schemata illustrated here, the task is to determine whether the model is an individual or a sociocultural one, or an amalgam.

34. It has unfortunately come to my – surprised – attention that because I notice so many racial incidences, some readers believe this shows that I must personally harbor racial prejudice. This perception disturbs me, not only because it is incorrect, but because it is logically faulty and detracts from the acceptance of my findings. The reasons I recognize so many incidences are as follows: First, being trained in social and cognitive psychology, I keep close track of this subject area. Second, growing up during times when racism was culturally widespread and generally acceptable, I was socialized into the full range of racial myths, images, and stereotypes; consequently, the schemata are stored in my memory. Third, having a best friend and colleague of thirty years who is black has sensitized me to a level of racial issues not typically possessed by whites. However, given the situated nature of knowledge I am aware

82 Discourse & Society 20(1)

that, in principle, there is the possibility that my work may reflect automatic activation of Racial1, Racial2, or Racial3 schemata based on my early conditioning. Attributions of prejudice – just as with all SubLit analyses in general – requires more than just an assertion; it requires a methodologically informed basis for the attri- bution. In the absence of methodologically informed evidence, to maintain that recognizing SubLit racial schemata reflects a personal prejudice or racism on my part is to say that all SubLit analyses reflect my personal schemata. This makes no sense.

R E F E R E N C E S

Baars, B.J., Cohen, J., Bower, G.H. and Berry, J.W. (1992) ‘Some Caveats on Testing the Freudian Slip Hypothesis: Problems in Systematic Replication’, in B.J. Baars (ed.) Experimental Slips and Human Error, pp. 289–313. New York: Plenum Press.

Bargh, J.A. (1997) ‘The Automaticity of Everyday Life’, in R.S. Wyer, Jr (ed.) The Automaticity of Everyday Life: Advances in Social Cognition, Vol. 10, pp. 1–61. Mahwah, NJ: Erlbaum.

Bargh, J.A. and Barndollar, K. (1996) ‘Automaticity in Action: The Unconscious as Repository of Chronic Goals and Motives’, in P.M. Gollwitzer and J.A. Bargh (eds) The Psychology of Action, pp. 457–71. New York: Guilford.

Bargh, J.A. and Gollwitzer, P.M. (1994) ‘Environmental Control of Goal-directed Action: Automatic and Strategic Contingencies Between Situations and Behavior’, in W.D. Spaulding (ed.) Nebraska Symposium on Motivation, Vol. 41, Integrative Views of Motiv- ation, Cognition, and Emotion, pp. 71–124. Lincoln, NE: University of Nebraska Press.

Bargh, J.A., Chen, M. and Burrows, L. (1996) ‘Automaticity of Social Behavior: Direct Effects of Trait Construct and Stereotype Activation on Action’, Journal of Personality and Social Psychology 71: 230–44.

Crosby, F., Bromley, S. and Saxe, L. (1980) ‘Recent Unobtrusive Studies of Black and White Discrimination and Prejudice: A Literature Review’, Psychological Bulletin 87: 456–563.

Devine, P.G. (1989) ‘Stereotypes and Prejudice: Their Automatic and Controlled Components’, Journal of Personality and Social Psychology 56: 680–90.

Dijksterhuis, A. and Nordgren, L.F. (2006) ‘A Theory of Unconscious Thought’, Perspectives on Psychological Science 1(2): 95–109.

Dijksterhuis, A. and Van Olden, Z. (2006) ‘On the Benefits of Thinking Unconsciously: Unconscious Thought Can Increase Post-Choice Satisfaction’, Journal of Experimental Social Psychology 42(5): 627–31.

Draine, S.C. and Greenwald, A.G. (1998) ‘Replicable Unconscious Semantic Priming’, Journal of Experimental Psychology: General 127: 286–303.

Duncan, B.L. (1976) ‘Dif ferential Social Perception and Attribution of Intergroup Violence: Testing the Lower Limits of Stereotyping of Blacks’, Journal of Personality and Social Psychology 34: 590–8.

Fromkin, V.A. (1973) Speech Errors as Linguistic Evidence. The Hague: Mouton. Gresson, A. (1982) The Dialectics of Betrayal: Sacrifice, Violation and the Oppressed. Norwood,

NJ: Ablex Pub. Gresson, A. (1995) The Recovery of Race in America. Minneapolis, MN: University of

Minnesota Press. Haskell, R.E. (1978) ‘An Analogic Model of Small Group Behavior’, International Journal

of Group Psychotherapy 28: 27–54.

Haskell: Unconscious linguistic referents to race 83

Haskell, R.E. (1982) ‘The Matrix of Group Talk: An Empirical Method of Analysis and Validation’, Small Group Behavior 13: 165–91.

Haskell, R.E. (1983) ‘Cognitive Structure and Transformation: An Empirical Model of the Psycholinguistic Function of Numbers in Discourse’, Small Group Behavior 14: 419–43.

Haskell, R.E. (1986–7) ‘Social Cognition, Language, and the Non-Conscious Expression of Racial Ideology’, Imagination, Cognition and Personality 6: 75–97.

Haskell, R.E. (1991) ‘An Analogical Methodology for Analysis and Validation of Anomalous Cognitive and Linguistic Operations in Small Group (Fantasy Theme) Reports’, Small Group Research 22: 443–74.

Haskell, R.E. (1999) Between the Lines: Unconscious Meaning in Everyday Conversation. New York: Plenum/Insight Books.

Haskell, R.E. (2001) Deep Listening: Uncovering Hidden Meaning in Conversations. Cambridge, MA: Perseus.

Haskell, R.E. (2002) ‘The New Cognitive Unconscious’, Evolution and Cognition 8(2): 184–99.

Haskell, R.E. (2003a) ‘A Logico-Mathematic, Structural Methodology: Part I, The Analysis and Validation of Sub-Literal (SubLit) Language and Cognition’, The Journal of Mind and Behavior 24: 347–400. Available at: http://faculty.une.edu/cas/rhaskell/HASKELL_ Method_Article_Part_I.pdf

Haskell, R.E. (2003b) ‘A Logico-Mathematic, Structural Methodology: Part II, Experimental Design and Epistemological Issues’, The Journal of Mind and Behavior 24: 401–22. Available at: http://faculty.une.edu/cas/rhaskell/HASKELL_Method_Article_Part_ II.pdf

Haskell, R.E. (2004) ‘A Logico-Mathematic, Structural Methodology: Part III, Theoretical, Evidential, and Corroborative Bases for Sub-Literal Cognition and Language’, Journal of Mind and Behavior 25: 287–322. Available at: http://faculty.une.edu/cas/rhaskell/ HASKELL_Method_Article_Part_III.pdf

Haskell, R.E. and Badalamenti, A. (2003) ‘Algebraic Structure of Verbal Narratives with Dual Meaning’, Mathematical and Computer Modelling 37: 383–93.

Hassin, R.R., Uleman, J.A. and Bargh, J.A. (eds) (2005) The New Unconscious. New York: Oxford University Press.

Kunde, W., Kiesel, A. and Hoffmann, J. (2003) ‘Conscious Control over the Content of Unconscious Cognition’, Cognition 88: 223–42.

Lewicki, P. (1986) Nonconscious Social Information Processing. New York: Academic Press. Merikle, P.M. (1998) ‘Psychological Investigations of Unconscious Perception’, Journal of

Consciousness Studies 5: 5–18. Available at: http://watarts.uwaterloo.ca/~pmerikle/ Montague, M.F.A. (1964) Man’s Most Dangerous Myth: The Fallacy of Race. New York:

Oxford University Press. Neisser, U. (1976) Cognition and Reality: Principals and Implications of Cognitive Psychology.

San Francisco, CA: W.H. Freeman. New York Times (1999) New England Edition, 29 January, p. A8. Pinker, S. (1999, February 2) ‘Racist Language, Real and Imagined’. New York Times. Reber, A.S. (1993) Implicit Learning and Tacit Knowledge: An Essay on the Cognitive

Unconscious. New York: Oxford University Press. Salgado, R.D. (2006) ‘Dan the Xenophobe Rides the A-Train or the Modern, “Secret”

Racist: Unconscious Racism in “Enlightened America” and the Destructive Stereotypes that Fuel It’. Working paper hosted by The Berkeley Electronic Press. ExpressO Preprint Series. Available at: http://law.bepress.com/cgi/viewcontent.cgi?article=4426&cont ext=expresso

84 Discourse & Society 20(1)

Sellen, A.J. and Norman, D. (1992) ‘The Psychology of Slips’, in B.J. Baars (ed.) Experimental Slips and Human Error, pp. 317–79. New York: Plenum Press.

Smith, D.L. (2004) Why We Lie: The Evolutionary Roots of Deception and the Unconscious Mind. New York: St. Martin’s Press.

Stemberger, J.P. (1992) ‘The Reliability and Replicability of Naturalistic Speech Error Data: A Comparison With Experimentally Induced Errors’, in B.J. Baars (ed.) Experimental Slips and Human Error, pp. 195–215. New York: Plenum Press.

Van Dijk, T.A. (1992) ‘Denial of Racism in Discourse’, Discourse & Society 3(1): 87–118. Van Dijk, T.A. (1997) ‘Cognitive Context Models and Discourse’, in M. Stamenow (ed.)

Language Structure, Discourse and the Access to Consciousness, pp. 189–226. Amsterdam: John Benjamins.

Van Dijk, T.A. (2002) ‘Discourse and Racism’, in D. Goldberg and J. Solomos (eds) The Blackwell Companion to Racial and Ethnic Studies, pp. 145–59. Oxford: Blackwell.

Van Dijk, T.A. (2006) ‘Discourse, Context and Cognition’, Discourse Studies 8(1): 159–77. Vanman, E.J., Paul, B.Y., Ito, T.A. and Miller, N. (1997) ‘The Modern Face of Prejudice

and Structural Features that Moderate the Effect of Cooperation on Affect’, Journal of Personality and Social Psychology 73: 941–59.

Velmans, M. (1991) ‘Is Human Information Processing Conscious?’, Behavioral and Brain Sciences 14: 651–726. Available at: http://cogprints.ecs.soton.ac.uk/archive/00000 593/00/199802003.html

Wittenbrink, B., Judd, C.M. and Park, B. (1997) ‘Evidence for Racial Prejudice at the Implicit Level and its Relationship with Questionnaire Measures’, Journal of Personality and Social Psychology 72: 262–74.

R O B E R T E . H A S K E L L is former Professor of psychology and department chair at the University of New England (USA), and is co-founder of The New England Institute of Cognitive Science and Evolutionary Psychology (www.une.edu/nei), which he co-directs. He earned his PhD from the Pennsylvania State University in Psychology and Social Relations; BA and MA from San Francisco State University. He is a Charter Member of the Association for Psychological Science. Major publications include: seven books, over 60 research articles, and invited entries for the Encyclopedia of Applied Psychology and the Academic American Encyclopedia. He also serves on several editorial review boards. Primary areas are Cognitive and Social Psychology; research areas include: Small Group Dynamics; Unconscious Cognition; Transfer of Learning, and Analogical Reasoning; the development of a Logico-mathematic Methodology. A D D R E S S : [email: [email protected] or [email protected]]