Todd Levon Brown, MPH, M.Arch, MA, M.Phil, Assoc. AIA; PhD Candidate in Environmental Psychology, City University of New York Graduate Center; Adjunct Faculty, Queens College, CUNY – Urban Studies Department, Fashion Institute of Technology, SUNY – Social Sciences Department 

While the role of architecture in (re)producing socioracial dichotomies has been a topic among scholars in recent years, it is still not fully understood as to what architectures are perceived as racialized or what psychosocial attributes and meanings are ascribed to such racialized readings of specific architectural aesthetics. This study seeks to understand how distinct architectures, in a gentrified neighborhood, are perceived as racialized and evaluated by different socio-racial groups. Thirty-four White and 78 non-White participants who lived, worked or attended school in East Harlem, New York were asked to evaluate images of pre- and post-gentrification architectural spaces in the neighborhood and to speculate on what ethnic groups they perceived to occupy these spaces. Most (52.9% - 85.7%) of the participants in both groups perceived older pre-gentrification architectural images as depicting non-White spaces and evaluated them negatively, while the opposite was true in regard to images of newer post-gentrification buildings and environments, where a majority(55.1% - 92.3%) of both participant groups reported perceiving these spaces as White and evaluated them positively. Forty percent of statistically significant evaluations were associated with the education or ethnicity of the participant. The findings of this research suggest that architecture itself can be understood through racial associations which can directly impact perceptions of the built environment. Most interestingly, the results suggest there are little differences overall across ethnic groups in how distinct architectures, when understood as being associated with a specific race, are both perceived and evaluated.

Keywords: architectural evaluation, racialized architecture, architectural meaning, architectural perception, race and space

Word Count: 5207

Introduction

The notion that architectural space¹ can both symbolize and embody social, racial or class constructs and be understood through psychological effect, via visual judgement or response, has been recognized by seemingly few scholars. One noted exception is urban geographer Neil Smith (1996) who most poignantly highlighted this concept in his commentary on the 1980s gentrification wars in New York City’s Lower East Side: “Several protestors used a police barricade to ram the glass-and-brass doors of the Christadora condominium…, which became a hated symbol of the neighborhood’s gentrification” (p. 4). Environmental psychologists and architectural evaluators Wolfgang Preiser and Jack Nasar (2008) state that “a key component of the social, psychological and cultural aspects of building performance, in fact the overall performance, includes building appearance, its evaluative quality, the meanings and evaluative response it conveys to the users” (p. 91). Preiser and Nasar go on to argue that understanding perceived visual quality in building performance evaluation is an important factor in creating more humane environments and that “meanings and aesthetics are not separate functions, [but that] appearances can draw people in or repel them” (p. 91). Nasar’s (1999) post-occupancy evaluation of the Wexner Centre for the Arts—designed by renowned architect Peter Eisenman as a facility to attract students—found that “many students never went inside, because they believed the staff was uppity” (Preiser & Nasar, 2008, p. 91). In this case, those students had never entered the building, but indicated that this meaning of superiority was expressed by the elements of the architectural design on the exterior, which inevitably disrupted its intended welcoming message. Both the Christadora condominiums and the Wexner Center examples are pertinent in aiding in the understanding that the aesthetic features of buildings and other designed spaces convey distinctive psychosocial meanings, perceived by both occupants and non-users who frequently interface with them.

Unfortunately, there has been very little research employed that investigates the emotive and connotative qualities of architectural space and aesthetics through the sociocultural influences of racial differences that exist both among the viewers and in the perception of the architectural spaces themselves. One early example of research that explored the role of racial and class differences in the dimension of architectural evaluation and meaning is the research of architect Henry Sanoff. His House Form and Preference (1970) study of low-income Blacks and middle-income White respondents sought to measure differences and variability in how these dichotomized groups assessed visual displays of ‘designed and non-designed house silhouettes” (p. 334). This project revealed considerable variance between the groups’ assessments and suggested that “background and environmental experiences influence preferences” (Sanoff, 1970, p. 334).

While this research proved to be insightful in demonstrating that differences in individuals’ socioracial backgrounds appear to be a factor in how they perceive architectural space, it was not designed to address what role, if any, differences in the perception of the racial characteristics of these spaces have as a factor in the evaluation and meaning of these spaces. Although empirical studies on the psychological evaluation of perceived racialized space are nearly non-existent, there has been considerable psychological research done in the area of racial biases and differences in the perception and evaluation of faces. Such studies have used differences in reaction time (Fazio, Jackson, Dunton & Williams, 1995) or amygdala activation (Phelps et al., 2000) to demonstrate variation in initial psychological responses of Whites viewing Black (or other minority) faces and vice versa. The authors of such studies consistently indicate that their design methodology was intended to reveal, more accurately, subtle racial behavior (biases, prejudices, and judgements) in evaluations among participants than their self-reported measures of racial prejudice (Musch & Klauer, 2003). The reasoning, as mentioned above, of such studies readily lends itself to the application of similar methodologies for understanding racial and other demographic influences in the perception and evaluation of aspects of physical spaces, including architecture and environmental aesthetics—e.g. the perception of racialized spaces. University of Cambridge geographer Susan Smith (1993) defines racialized space as “the process by which residential location is taken as an index of the attitudes, values, behavioural inclinations and social norms of the kinds of people who are assumed to live [there]” (Calmore, 1995, p.1236).

The primary goal of this research is to investigate whether different ethnic and social groups within a single neighborhood—in this case, a gentrifying neighborhood in New York City—might perceive certain architectures and built environments as racialized and how this contributes to their evaluation of such spaces. An additional aim is to investigate how architectural space might embody socioracial differentiation and potentially create different experiences of desirability, particularly in socio-spatially hostile environments such as gentrified communities. Therefore, this research explores three questions, the first two of which are interconnected: 1) Is architecture perceived as racialized? 2) How does individuals’ perception of architectural space as being racialized affect their evaluation of the space? 3) Do demographic differences between groups, such as ethnicity and socioeconomic status, correlate with differences in how they evaluate, positively or negatively, architecture and how they assign ethnic attributes to the built environment (buildings, streetscapes, etc)?

Methodology

Participants in this study consisted of 112 male (n = 50) and female (n = 62) residents, students or employees in East Harlem, Manhattan, New York between the ages of 18 and 72 (M_age = 32.05, SD = 13.42). Participant ethnicity demographics were as follows: 25.9% Black (n = 29), 21.4% White (n = 24), 27.7% non-White Hispanic (n = 31), 8.9 % White Hispanic (n = 10), 3.6% Asian (n = 4), and 12.5% multi-racial (n = 14). Combining participants self-identified as White or White( Hispanic) into a ‘White’ ethnic category and all other participants into a ‘non-White’ minority category, the ethnicity demographics were 30.4% White (n = 34) and 69.6 % non-White (n = 78).

This project employed the use of six images of architectural details and spaces all taken within a 4-block radius of 116^th Street and Lexington Avenue within the East Harlem community. The photos captured three different spatial scales and architectural aesthetic conditions: (1) a residential doorway, (2) a complete residential building façade and (3) a mixed-use streetscape. These six images were divided into two groups: (1) post-gentrification or new construction (less than 10 years old) and (2) pre-gentrification or vernacular neighborhood architecture (older than 30 years). The categorization of these images as pre- or post-gentrification was created by the researcher and was not known to participants at the time of the study; however, it was assumed that participants were familiar with at least some of the buildings depicted in the images, given their status as residents, students or employees in the neighborhood. Each of the architectural photographs was accompanied by a 13-item, five-point semantic differential scale containing bipolar descriptive adjectives to be used for participants’ perception and evaluation of the images. The 13 word pairs used were extracted from Sanoff’s (1970) House Form and Preference study which consisted of 19 bipolar word pairs, and were chosen because the researcher deemed them the least ambiguous and also as having the most potential to connote positive and negative meanings and attributes. The 13 bipolar word pairs used in the semantic differential scale were as follows: “welcoming/unwelcoming”, “pleasant/unpleasant”, “relaxed/tense”, “rich/poor”, “positive/negative”, “interesting/dull”, “happy/sad”, “like/dislike”, “beautiful/ugly”, “safe/unsafe”, “exclusive/inclusive”, “clean/dirty” and “peaceful/fearful”. The first word in each of these pairs was assumed to connote a positive descriptor while the second word in the pair was assumed to denote a negative evaluation. The ordering of the ‘negative’ and ‘positive’ words was randomly alternated so that not all ‘positive’ adjectives occurred on the same side of the scales to reduce response bias. After participants evaluated the six images using the semantic differential scale, they were shown the same six images again, in a different order. This time, participants were first asked to select what ethnic group(s) they perceived to live in or occupy the architectural spaces represented in the photographs. They were given six ethnic categories to choose from: Hispanic, White, Black, Asian, and Middle Eastern as well as an ‘Other’ designation. In the second question, participants were then asked to select what they perceived to be the majority or primary ethnic group(s) in these spaces from the above categories, or they could select that they did not perceive any specific ethnic group to be the primary occupant of the space represented in the image. Lastly, at the end of the survey, participants were asked to provide demographic information. In order for any architectural space depicted in one of the survey photographs to be classified as being perceived as White, the participant must have selected only the “White” option out of all six possible ethnic categories to choose from in the first question. Architectural images classified as primarily White meant that the participant selected only the “White” option in the second question about the perceived majority or primary ethnic groups occupying the space depicted in the image, even if they selected multiple ethnic groups (including White) for the first question. The opposite is true for any image classified as being perceived as depicting non-White space. In this case, a participant must have selected any ethnic group(s) except “White” when answering the first question on their perception of the ethnic group(s) occupying the space depicted in the photograph. Similarly, the image was categorized as being perceived primarily as depicting a non-White space if the participant selected any ethnic group(s) except “White” when answering the second question about the perceived majority or primary ethnic groups occupying the space depicted in the image, even if they selected “White” in the first question.

This research was conducted using both paper (n = 62) and online (n = 50) surveys, administered to participants who identified themselves as residents, students or employees within the East Harlem Community of Manhattan, New York. Recruitment sites included residential buildings, retail shops, barbershops, community centers, and educational institutions.

The data were analyzed using IBM SPSS software versions 20 and 22. The scoring for the 5-point semantic differential scale was transformed from a 1-5 scale to a -2 to +2 scale so that average values ranging from -2 to 0 represented below-the-midpoint, or ‘negative’, evaluations corresponding to the assumed ‘negative’ words in the word pairs. Average values of 0.00001 to 2 represented a correlation to the operationally defined ‘positive’ words in the word pairs and therefore were categorized as ‘positive’, or above-the-midpoint, evaluations.

Results

The results represent an exploratory analysis of the data that sought to discover relationships between the evaluation of architecture (per photos via the semantic differential scale), the perception of which ethnic group(s) resided in these spaces, and the socioracial demographics of the study participants. Additionally, trends, patterns, frequencies and distributions of the data are reported that allow for an understanding of which, if any, variables influenced differences in participants’ evaluation and perception of the architectures depicted in the images.

Table 1 lists the percentages of both White and non-White participants who perceived each image as a ‘White’ (being primarily occupied by or associated with White people) or ‘non-White’ space. The table also lists the percentages, for both groups, of those who evaluated each image ‘positively’ and ‘negatively’. For each of the older, pre-gentrification images, most (52.9% - 85.7%) of both participant groups perceived these to be ‘non-White’ spaces. Also, for each of the newer, post-gentrification architectural photographs, a majority (55.1% - 92.3%) of both groups perceived them as ‘White’ spaces. With the regard to perceptions of the primary ethnic groups perceived to be associated with the spaces depicted in these photos, results from study participants follow the same trend, except in the case of the image of the post-gentrification streetscape. Here, roughly two-thirds of both groups perceived this photo to be that of a primarily ‘non-White’ space (See Table 1).

The individual images also received different mean scoring evaluations based on tabulated averages of their accompanying bipolar scale. All images depicting post-gentrification (newer) architectural spaces received positive mean scores for each of the 13 word pairs, with one exception: the image of the post-gentrification streetscape received a negative mean score (-0.29) for the ‘exclusive/inclusive’ word pair (See Table 2). Additionally, every image depicting a pre-gentrification (older) architectural space received negative mean scores for all 13 word pairs, with one exception: the image of pre-gentrification façade received a positive sample mean score (0.07) for the ‘relaxed/tense’ word pair (See Table 2).

Highest and lowest mean scores for each of the 13 word pairs occurred for post-gentrification and pre-gentrification images, respectively. Furthermore, 12 of the 13 lowest mean scores (92.3%) were associated with the image of the pre-gentrification doorway. For each of the 13 word pairs, the highest mean scores—and thus, greatest ‘positive’ evaluation—were associated with one of the post-gentrification images (See Table 2).

To determine which demographic variables demonstrated the strongest relationship to participant responses for word pairs items, chi-square tests for independence were conducted that analyzed cross-tab relationships between the 13 word pairs for each of the six (6) images and nine (9) demographic variables: ethnicity, education, gender, income, political views (liberal - conservative spectrum), gentrification views (as being positive or negative), years of residence in East Harlem, residential identity (participants’ identification of their residential status in East Harlem), and family status (participants’ report of their family’s residential status in East Harlem). This approach resulted in 702 distinct chi-square analyses, of which 107 proved to be statistically significant. The 107 statistically significant chi-square tests revealed that education and ethnicity were most frequently associated with participants’ responses on the semantic differential scale, representing 24 and 19 of the tests, respectively. 22.4% (n = 24) of the significant chi-square tests were due to education and 17.75% (n = 19) were the result of participants’ ethnicity.

Discussion

This study presents three hypotheses: 1) that individuals will perceive architectural space as racialized, 2) that individuals’ perception of architecture as racialized correlates to how they evaluate it as ‘positive’ or ‘negative’ and 3) that groups of individuals who vary by ethnicity or socioeconomic status will differ in how they evaluate different types of architectural spaces perceived as racialized. The first hypothesis specifically assumed that the images of the older, pre-gentrification architectural spaces would be perceived as primarily ‘non-White’ spaces and conversely that the newer architectures would be perceived as ‘White’ spaces. The second hypothesis more specifically assumed that older, pre-gentrification architectural spaces, perceived as ‘non -White’, would also be evaluated as ‘negative’ spaces and that newer, post-gentrification spaces, perceived as ‘White’, would also be evaluated as ‘positive’ spaces. The third hypothesis more specifically assumed that in the contrast of older (pre-gentrification) architectural spaces and newer (post-gentrification) spaces, the non-White participant group would be more likely to evaluate images of pre-gentrification architecture more positively than their White counterparts and the White participants would be more likely to provide a more positive evaluation for post-gentrification spaces than their non-White counterparts. The findings of this study supported the first two hypotheses. However, the third hypothesis was not only rejected by this research, but the findings seem to imply that the exact opposite may be true.

Concerning the first two research questions and corresponding hypotheses, the data suggest that individuals indeed ascribe racial attributes to architecture itself and that the perception of architectural space as racialized corresponded to how study participants evaluated the spaces as ‘positive’ or ‘negative’, via the semantic differential bipolar word pairs. A majority of research participants, in both ethnic groups, perceived all of the images depicting pre-gentrification, or older, architectural spaces to be both generally and primarily occupied by non-Whites (See Table 1 and Figures 1 & 2). The survey responses for these images clearly indicate that most of the participants did not perceive Whites to be at least one of the general or primary ethnic groups associated with the older architectural spaces. Additionally, participants in the sample were consistent in their overall evaluation of these images as ‘negative’ (See Table 1 and Figures 1 & 2). Conversely, a majority of research participants, in both ethnic groups, perceived Whites to be the only general or primary occupants of the newer, post-gentrification architectural spaces as represented by the images. The vast majority of survey respondents similarly shared this perception and also were consistent in their overall evaluation of the images of newer spaces as ‘positive’, as defined by the association of these images with the ‘positive’ words in the bipolar pairs (See Table 1 and Figures 1 & 2).

The second hypothesis yielded interesting, yet counterintuitive results. There was virtually no statistically significant difference in how White and non-White participants perceived and evaluated the architectural images categorically. The vast majority of both groups shared similar perceptions as to the racial groups they associated with the images of the pre-gentrification and post-gentrification architectural spaces (See Table 1). Most members of each group were also consistent in perceiving pre-gentrification architectural images as depicting ‘non-White’ spaces and evaluated them ‘negatively’ overall, while images of post-gentrification architectural spaces were perceived as being occupied by Whites and were evaluated ‘positively’ overall by both groups. However, a between-groups visual analysis of evaluation patterns interestingly reveals that, overall, Whites evaluated the older, pre-gentrification images more positively than non-Whites, except in the case of the pre-gentrification commercial streetscape (See Figure 1). As for the newer, post-gentrification images, a between-groups visual analysis shows that non-Whites evaluated these images similarly or more ‘positively’ than Whites (See Figure 1). Overall, five of the six images produced mean evaluation results, by participants’ categorical ethnic groups, that were different or opposite to what was expected in the third research hypothesis.

The results of this study appear to indicate that different types of architectural spaces do embody differentiated social and racial perceptions as well as varied meanings via the interpretation of their aesthetic features. While the study was not designed to indicate what specific elements in the architectural photographs the participants were responding to, the findings of this research strongly support consistent patterns of perceiving newer architectural spaces as ‘positive’ ‘White’ milieus and older pre-gentrification architectural spaces as ‘negative’ ‘non-White’ environments. One explanation of this consistency across ethnic, and particularly class, groups may be contextualized by the seminal work of French sociologist and philosopher Pierre Bourdieu. In his 1979 book, Distinction: A Social Critique of the Judgement of Taste, he proposes that those in possession of a “high volume of cultural capital” (e.g. non-financial social assets—such as education, style of speech, style of dress, etc.—which promote social mobility beyond economic means) are also those who are most likely to be able to determine what constitutes taste within society. Along with taste, the distinction of what is high and low culture, is also readily accepted by these disenfranchised masses as legitimate and natural, which inevitably and concurrently causes them to yield to existing restrictions on conversions between the various forms of capital (economic, social, cultural). Applied to architectural spaces—as physical and environmental capital, Bourdieu’s framework elucidates how hegemonic constructions and perceptions of what constitutes ‘elite/superior/good’ and ‘common/inferior/bad’ spatial environments permeate the whole of society. These notions are further perpetuated and reproduced via avenues such as the mass media, which is also controlled by a select elite (Dye, 2014). Relationships between racial groups, social classes and types of living environments are reinforced to the masses via television programs such as MTV’s Cribs, Million Dollar Listings, and almost every reality show: they consistently provide the public with an image of what good and successful living looks like. Even when successful non-Whites are shown living in affluence, usually it is because they are perceived as living in a White, affluent community. Additionally, the fact remains that Whites sit atop the socioeconomic ladder in the US (and almost every other part of the world) and that in almost every urban area, Whites are generally associated with affluence and non-Whites with poverty (Larson, 2011; Wildman, Armstrong & Moran, 2012). Using this framework possibly explains why this research demonstrated a significant consistency in how participants across ethnic and socioeconomic groups evaluated these spaces.

In the case of non-White individuals who perceive the environments in which non-Whites live as ‘negative’ spaces, one must wonder how this understanding plays a role in the psychosocial relationship that is constructed and reproduced daily in the context of a gentrifying neighborhood. One’s perception and cognition of the environment in which one lives can be central to that individual’s formation and understanding of his or her own identity. Bush, Moffat and Dunn’s (2001) grounded research on place, health and stigma builds on sociologist Erving Goffman’s (1963) foundational work on how the mechanism of social stigma can produce a “spoiled identity” in an individual. Bush and colleagues applied this theory to environmental mechanisms as well and found that those who live in environmentally compromised spaces can begin to develop a “spoiled identity” as they internalize all of the stigmas attributed to their physical domain. Wildman, Moran and Armstrong (2012) argue that “Whites tend not to notice that they too have a race that carries social meaning and generally positive presumptions” (p. 1016). Non-Whites, not being afforded such privilege, are constantly forced to deal with their socioracial status as ‘other’ and often clearly realize that their race—and the places they live—also carry social meaning and generally negative presumptions. As the results of this study show that non-Whites in East Harlem view the pre-gentrification residential spaces in which they live as ‘negative’ or inferior to the architectures of gentrification, many questions arise about how the identities of these minority groups might become further spoiled as well. In neighborhoods such as East Harlem, New York, the juxtaposition of these architectures of gentrification, in contrast with perceived spaces of despair, can serve to reinforce the perceptions of inferiority that plague these physical places and may be ascribed to residents of the community themselves. The resultant environmentally differentiated and reinforced identities produced between White and non-White ethnic groups in the socio-economically and spatially hostile environment of gentrification can serve as a mechanism for further polarization. Those residing in geographies socially constructed and psychologically perceived as marginalized may respond through acts of despair or resistance (Smith, 1996; Yiftachel, 2009), both of which leave little room for reconciliation with their gentrifying neighbors.

Limitations

There are several limitations to this research that should be noted. The most important is the use of the primary measurement tool for analysis: the semantic differential scale itself. Use of the semantic differential scales has incurred criticism from several scholars (Infante, 1980). Such scholars have identified numerous problems that emerge when using bipolar rating scales: concept-scale interaction, rater-scale interaction, source dimensions that are an artifact of factoring procedures, and haphazard scale selection and factor-naming procedures (Cronkhite and Liska, 1976). The developers of the scale, psychologist Charles Osgood and colleagues, found that there are three recurring attitudes that people use to assess words and phrases: evaluation, potency, and activity (Osgood, Suci and Tannenbaum, 1957; Osgood, May and Miron, 1975). They also found that the evaluative factor accounted for most of the variance in scaling and related this to the idea of attitudes (Himmelfarb, 1993). Himmelfarb (1993) also noted that one major problem with this scale is that its psychometric properties and level of measurement are disputed. This strongly suggests that nature of information associated with the ratings on the scale are not markedly clear due to the highly subjective nature of bipolar adjective pairs. What the actual meaning of each word in the pair is to each user of the scale may present a considerable level of variability (Kirch, 2008). However, in this study, observation of the distributions and central tendencies of participants’ responses on the bipolar scale indicated consistency and agreement among most of the participants in both ethnic groups (See Figure 2).

A discussion of the limitations of this research must also consider the sample of respondents who participated in the study, along with their categorization and size of the sample. A critical aspect of this study was the comparison of responses from White and non-White participants. Although other quantitative research has used the White/non-White categories for participant race/ethnicity (Barnshaw & Letukas, 2010), such binaries with aggregated race variables often do not allow for “the investigation of differences among and between visible minority groups”, and often obfuscate the role of racial identity intersectionality with socio-economic variables such as education or household income (Rouhani, 2014, p. 3). It is also conceivable that differences might emerge within the ‘White’ group as well, across those who identify as White verses White Hispanic. The aggregated binary ethnic classification of the participants proved to be useful for the statistical analysis of such a small sample of 112 respondents. However, such a small sample did not permit considerable further exploration of such intersectionality to explore responses within the two ethnic groups and across different categories of education and income. For example, a chi-square analysis (X² (1, N = 112) = 19.334, p = .000) of the ethnicity and education variables revealed that they were highly correlated, r(111) = -.415, p = .000. The education variable was also dichotomized to represent two groups: those with a high school diploma or less and those with a college degree or higher. The two variables ethnicity and education are highly conflated, making it difficult to determine if the participants’ race or class, or an interaction of the two is a greater predictor of perception and evaluation.

Lastly, the design of some of the survey questions themselves are important in understanding the limitations of this study. Respondents were given the option to select multiple choices for identifying what ethnic groups they perceived to ‘live in’ each space, as well as for which ethnic groups they perceived to be ‘the primary occupants’ of the architectural spaces represented by the images. In many instances, allowing respondents to select from more than one ethnic category did not allow the research to conclusively categorize participants’ perceptions of a space as being ‘White’ or ‘non-White’. This means that for many cases, only the absence of selecting White was the determinant factor for categorizing the perception of a space as ‘non-White’. In cases where “White” was selected along with other ethnic groups, it is impossible to know how the space was predominantly perceived. Future studies along the same line of inquiry might better determine how participants predominantly perceive the ethnic orientation, or racialization, of a space by soliciting a single response, at least for the perception of primary ethnic group occupancy. However, despite all these limitations, it is still true that the research produced consistent differences in participants’ perceptions of the images of pre-gentrification and post-gentrification architectural spaces. This is therefore a promising direction in understanding how these spaces are perceived as racialized.

Conclusion

This study suggests that people do indeed perceive and understand the physical environment, including architecture, as racialized. Racial biases, stereotypes and assumptions are ascribed to physical and architectural spaces, just as they are to other socio-cultural phenomena (e.g. food, music, clothing, etc.). Individuals’ perceptions of physical spaces as racialized appear to be correlated to their judgements and evaluations of these spaces. When physical spaces are perceived as being primarily associated with White people, these spaces are overall evaluated as desirable and ‘positive’ areas; those associated primarily with non-Whites are deemed undesirable and ‘negative’ environments overall. Contrary to one of the hypothesized outcomes, in this study there appears to be no difference across ethnic categorical groups in how specific architectural spaces are perceived and evaluated. Cross-racial meanings of what constitutes good and bad architecture, while possibly dictated by a select elite, seems to have been adopted as a shared western cultural standard, and this in turn appears to be the primary determinant for these consistencies. Although this research demonstrates a connection between the racialized perception of architectural space and how it is evaluated, it does not provide an understanding of what architectural elements were being evaluated, or why, and leaves much to speculation. Further qualitative studies along this vein of research should allow for an in-depth contextual analysis of individuals’ perception of and relation to different architectural spaces in gentrifying communities.

References

Barnshaw, J., & Letukas, L. (2010). The low down on the down low: Origins, risk identification

and intervention. Health Sociology Review, 19(4), 478-490.

Bourdieu, P. (1984). Distinctions. A Social Critique of the Judgement of Taste. (R. Nice, Trans.). Cambridge, MA: Harvard University Press (Original work published 1979).

Bush, J., Moffatt, S. & Dunn, C. (2001). ‘Even the birds round here cough’: stigma, air pollution

and health in Teesside. Health & Place,7, 47 – 56.

Calmore, J. (1995). Racialized Space and the Culture of Segregation: "Hewing a Stone of Hope from a Mountain of Despair". University of Pennsylvania Law Review, 143(5), 1233-1273. doi:10.2307/3312475

Cronkhite, G. & Liska, J. (1976). A Critique of Factor Analytic Approaches to the Study of
Credibility. Communication Monographs, 43(2), 91-107.

Dye, T.R. (2014). Who’s Running America? The Obama Reign. 8^th ed. New York: Routledge

Fazio, R; Jackson, J.R.; Dunton, B.C.; & Williams, C.J. (1995). ‘Variability in automatic activation as an unobtrusive measure of racial attitudes: A bona fide pipeline?’. Journal of Personality and Social Psychology, 69, 1013–1027.

Goffman, E. (1963). Stigma: Notes on the Management of Spoiled Identity. Englewood Cliffs, NJ: Prentice-Hall.

Himmelfarb, S. (1993). The measurement of attitudes. In A.H. Eagly & S. Chaiken (Eds.), Psychology of Attitudes, 23-88. Thomson/Wadsworth

Infante, D.A. (1980). The Construct Validity of Semantic Differential Scales for the Measurement of Source Credibility. Communication Quarterly, 28(2), 19-26

Kirch, W. (Ed.). (2008). Level of measurement. Encyclopedia of Public Health 2. Springer. pp. 851–852.

Larson, D. (2011). Perceptions of Whiteness and Class Privilege in a Post-Obama America. (Master’s thesis). Retrieved from http://preserve.lehigh.edu/etd. Paper 1283.

Musch, J., & Klauer, K.C. (Eds.). (2003). The Psychology of Evaluation: Affective Processes in Cognition and Emotion. Mahwah, NJ: Lawrence Erlbaum Associates.

Nasar, J.L. (1999). Design by Competition: Making Design Competition Work. New York: Cambridge University Press.

Osgood, C. E., May, W. H., & Miron, M. S. (1975). Cross-Cultural Universals of Affective Meaning. Urbana, IL: University of Illinois Press

Osgood, C.E., Suci, G., & Tannenbaum, P. (1957). The measurement of meaning. Urbana, IL: University of Illinois Press

Phelps, E.A., O’Connor, K.J., Cunningham, W.A., Funayama, E.S., Gatenby, C., Gore, J.C., Banaji, M. (2000). Performance on Indirect Measures of Race Evaluation Predicts Amygdala Activity. Journal of Cognitive Neuroscience, 12(5), 729-738.

Preiser, W. & Nasar, J. (2008). Assessing Building Performance: Its Evolution from Post-Occupancy Evaluation. Archnet-IJAR, International Journal of Architectural Research, 2(1), 84-99.

Rouhani, S. (2014). Intersectionality-informed Quantitative Research: A Primer. The Institute for Intersectionality Research and Policy, SFU. April 2014.

Smith, N. (1996). ’Class Struggle on Avenue B: The Lower East Side as Wild Wild West’ in

The New Urban Frontier: Gentrification and the Revanchist City. New York: Routledge.

Smith, S. (1993). Residential Segregation and the Politics of Racialization. In M. Cross & M. Keiths (Eds.), Racism, The City And The State (pp.128 - 144). New York: Routledge.

Sanoff, H. (1970). House Form and Preference. Paper presented at EDRA 2: Proceedings of the 2^nd Annual Environmental Design Research Association Conference, Pitsburgh. Pp. 334-339.

Wildman, S.M., Armsrong, M., & Moran, B. (2012). Revisiting the Work We Know So Little About: Race, Wealth, Privilege and Social Justice. UC Irvine Law Review, 2, 1011-1022.

Yiftachel, O. (2009). Critical theory and ‘gray space’: Mobilization of the colonized. City, 13(2-3), 240-256.

Tables

Table 1

Percentages of Perceived Ethnicity by Image Type and Participant Ethnicity

^a The categories (‘White’ and ‘non-White’) denote when participants indicated that they perceived Whites to be at least one of the ethnic groups living in the space depicted by the image. ‘Non-White’ percentages indicate the rate of participants who did not select White as at least one of the ethnic groups they perceived to live in the space. The categories (‘White’ and ‘non-White’) for ‘Perceived Primary Ethnicity’ are the same as above but denote when survey participants indicated that they perceived Whites to be at least one of the primary occupants of the space.

Table 2

Mean Values for Semantic Differential Word Pairs by Image Type

Figures

Figure 1. Side-by-side visual comparison of average semantic differential scores for bipolar word pairs by ethnicity, including overall sample means, for post-gentrification façade (left) and pre-gentrification doorway(right)..

*Note: these two images are representative of the fact that images depicting newer, post-gentrification architectural spaces received positive mean evaluation scores while the opposite is true of images depicting older, pre-gentrification spaces. The darker line represents the scores from non-White participants, the lighter line those of White participants and the dashed line is the mean of the total sample.

Figure 2. Histograms of participant evaluation scores by image type by participant ethnicity.