Desert dreamscapes: Residential landscape preference and behavior

https://doi.org/10.1016/j.landurbplan.2005.06.002Get rights and content

Abstract

The residential landscape constitutes a significant portion of the urban environment. With the increasing mobility of our society, many people come to reside in environments with unfamiliar plant communities and environmental conditions. In this research, 232 Phoenix, Arizona homeowners were surveyed to investigate their residential landscape preferences and to what degree these preferences were reflected in their actual behaviors in their front and backyards. Landscape preferences vary between the front and backyard residential landscapes in a manner consistent with [Goffman, E., 1959. The Presentation of Self in Everyday Life. Anchor Books, New York] about the symbolic presentation of self. When a multinomial logistic regression model was constructed to predict front yard landscape preference with the independent variables income, length of residence in the Phoenix area, degree of environmental concern and engagement in desert recreational activities, income was the only significant predictor. In the front yard, lower-income homeowners tend to prefer lawn, middle-income homeowners preferred desert landscaping and higher-income homeowners’ preferences were divided between desert and oasis landscape. While backyard landscape preferences varied among income groups, income was not a significant predictor of backyard landscape preference. Landscape preferences for the front and backyard were significantly different and 55% of respondents preferred different landscape typologies for the front and backyards. Overall, for landscape preference in the front yard, form follows class-specific fashion. For landscape preference in the backyard, form is more likely to follow individual fantasy. Landscape preferences were then compared with landscape behaviors. One-third of respondents expressed landscape preferences that were different from their landscape behaviors. However, the development industry has a significant influence on the design and construction of residential neighborhoods in Phoenix [Kirby, A., 2000. All new, improved! Cities 17(1), 1–5]. In an effort to sell homes, developers’ anticipate homeowners’ tastes and package their homes with desirable front yard ‘dreamscapes’ [Jencks, C., 1993. Heteropolis: Los Angeles, the Riots and the Strange Beauty of Hetero-Architecture, Academ Editions, London]. Therefore, we hypothesized that the macro-level influence of the developer's landscape legacy would have a greater impact on the front yard's appearance (behavior) than on the backyard's appearance. For the front yard, both the legacy and the homeowner's preference were significant predictors of landscape behavior, but in the backyard, only the homeowner's preference had a significant influence. The manifestations of the residential landscape reflect expressions of self, status, and conceptions of place that combine to create little understood ‘dreamscapes’. We must recognize the importance of the front yard as a visible symbol of self, the backyard's role as a personal pleasure ground, and the conflict that may occur when the natural landscape is relatively inhospitable to domestic behaviors.

Introduction

The residential landscape is a repository for incompatible ideologies. Since the single-family house in the United States is private property, it embodies the Jeffersonian ideals of individual freedom, democratic participation and unconstrained opportunity. However, the reality of the residential landscape immediately challenges these ideals, which are ingrained not only in the United States, but throughout many parts of the capitalist world. Economic sorting spatially clusters ‘like’ residents by income and social pressures encourage conformity. Today, most homeowners view their home as an important financial investment and their desire to eventually profit from its resale also influences their attitudes and behaviors. Further complicating the meaning of the residential landscape is its connection to the ‘natural’ landscape and, perhaps most importantly, its symbolic representation of the homeowner.

The increasing mobility of our society also has implications for the residential landscape as many people now find themselves living in regions that have very different plant communities. In Phoenix, Arizona, the majority of homeowners have moved there from other locations within the United States and Mexico and in this Sonoran desert ecosystem, which is unfamiliar to many migrants, landscape form often appears to follow fantasy. As a city of suburban character and exponential growth, four common landscape typologies embody different residential landscape ideologies. Each landscape typology serves as a distinct ‘dreamscape’. Jencks (1993, p. 73), defined dreamscapes as “marketable and commodified landscapes designed to satisfy fantasies of urban history”.

In an effort to understand landscape preferences in Phoenix, Arizona we explore the ideas of sociologists Goffman (1959) and Veblen (1899). These ideas lead us to hypothesize that residential landscapes are presentations of social class. We also hypothesize that preference for the different landscapes will depend upon a homeowner's length of residence in the Phoenix area, degree of environmental concern, and engagement in desert recreational activities.

The second section of this paper examines how closely landscape preferences reflect landscape behavior. Although we assume that people will make changes to their landscapes to reflect their preferences, new and older homes generally come with established landscapes and many owners opt to accept these landscape legacies. Increasingly, large developers dominate the design and construction of residential neighborhoods in Phoenix (Kirby et al., 2003). In an effort to sell new homes and provide visually consistent streetscapes, residential landscapes generally reflect the developer's choices. The developer's choice of the initial landscape composition functions as a macro-level force. Therefore, we hypothesize that the developer's choices, as reflected by the year of construction, and the homeowner's landscape preferences will influence landscape behavior.

Although much research has focused on the management of publicly held landscapes, the residential landscape has received less attention despite the fact that it occupies a significant proportion of the urban and suburban environment. When partitioning the greenspace in Syracuse New York, Richards et al. (1984) discovered that public street sides and public parks comprised 7 and 9%, respectively, while residential yards constituted 48%. The importance of the residential landscape was further confirmed by an Edinburgh, Scotland tree inventory (Last et al., 1976). Eighty-four percent of the city's trees were within residential lots.

Understanding how homeowners shape and maintain their private landscapes has important ecological implications. Within a desert environment, the treatment of the residential landscape has significant implications for water consumption, land subsidence due to overpumping groundwater, interference with surface hydrology and flood pathways, soil erosion (dust production and storage) and climate change (Warren et al., 1996, Martin, 2001, Stefanov et al., 2004).

In an effort to understand the residential landscape within a larger societal context, we first draw upon the work of Goffman and Veblen. The residential landscape, as the transition area around the domestic home and its neighborhood, has physical as well as psychological dimensions. Within American culture, the expectations for appearance and function vary depending on whether one is discussing the front or backyard. While this distinction can be anecdotally observed, much of the existing research on residential landscapes focuses on the visible front yard.

Goffman's concept of the presentation of self provides a framework by which to understand why front and backyards may differ in appearance. Goffman (1959) theorized that individuals would cultivate particular behaviors and shape their environments in an effort to communicate social status and identity. An individual's image manipulation may mingle between conscious awareness and unconscious emulation. Goffman's concept of the presentation of self seems consistent with Nassauer (1995) findings. In an investigation of why natural habitat areas are resisted by many American homeowners, Nassauer noted that the ‘messy’ appearance of ecologically valuable areas contradicted the social prescribed expectations for Midwestern stewardship. The agrarian tradition of the Midwest in part underlay the expectation that homeowners will maintain neat and orderly yards consistent with the practices of good farming.

Goffman also introduced the terms ‘front stage’ and ‘back stage’ as metaphors to explain differences in behavior. On the highly visible front stage, we are more conscious of being seen and we are more careful about the behaviors we undertake. On the ‘back stage’, we perceive less visibility and our behavior is less constrained. Goffman applied this concept of differentiated spaces within the house but, in his 1959 book, The Presentation of Self in Everyday Life, he made only one reference to the house's exterior. He wrote “in all classes in our society there is a tendency to make a division between the front and back parts of residential exteriors. The front tends to be relatively well decorated, well repaired and tidy; the rear tends to be relatively unprepossessing” (Goffman, 1959, p. 123).

If we apply Goffman's front stage and back stage metaphor to the residential landscape, the visible front yard is the front stage. In the front yard, the house and its landscape combine to project an image of the homeowner that is constantly on public display. The less visible backyard landscape, experienced by only invited guests and the home's occupants, more closely embodies Goffman's concept of back stage. The appearance and management of the backyard is more likely to deviate from social conventions and may prioritize the user's comfort over public impression. Therefore, we hypothesize that front and backyards will have very different appearances.

We believe that another key determinant of landscape preference and behavior is social class, which we are equating with income. Goffman's notion of the importance of class in shaping tastes or preferences is based upon the earlier work of Veblen. In 1894, Veblen wrote an article for The Popular Science Monthly that proposed an explanation of women's fashions. Veblen (1899, p. 85) believed that much of fashion was linked with the display of ‘conspicuous consumption’ and that fashions served as important symbols of ourselves. He advanced this idea in a book entitled, The Theory of the Leisure Class, and wrote “no class of society, even the most abjectly poor, foregoes all customary conspicuous consumption”. He stated that, next in importance to the instinct for self-preservation, was the instinct for emulation. Veblen believed that desirable tastes were established by a society's wealthiest members and “as specific consumption usage become conventional within the leisure class, self conscious status competition assures their imitation, to the extent possible, down the social ladder” (Brown, 1998, p. 9).

Duncan (1973, p. 355) wrote that “landscape tastes have been overlooked by students of social stratification. Yet they are a critical part of presentation of self for middle and upper class Americans, whose social interactions take place to a great extent in the home landscape”. Duncan identified alpha and beta landscapes in Westchester County, New York. The alpha landscape, favored by the more affluent homeowners, featured the naturalistic irregularity of the picturesque English landscape. Duncan believed that this contrived appearance of irregularity and picturesque opportunities was an expression of the homeowner's desires for privacy, solitude, and an effort to down-play affluence. The beta landscape of Westchester County featured open expanses of grass and symmetrical arrangements of trees and shrubs. The crisp organization of the beta landscape allowed for a clear view of the house from the road and, in Duncan's opinion, was more ‘American” then the alpha English landscape. Marcus (1995) has also noted the ego-symbolism of the home.

The landscape preference literature also contains insights that could improve our ability to predict which Phoenix residents will opt for which landscape typology. Within this extensive literature, some consistent patterns arise. Cross-cultural comparisons have found a surprising degree of consistency across the preferences of respondents (Berlyne, 1974, Ulrich, 1977, Ulrich, 1986, Brierley Newell, 1997). Cross-cultural similarities suggest that familiarity with a landscape does not necessarily contribute to preference. Kaplan and Herbert (1987, p. 291) found that “while preference is affected by familiarity, people do not necessarily prefer what they are familiar with”. Kennedy and Zube (1991) explored this notion by surveying residents of Tucson, Arizona and students attending the University of Arizona to determine their preferences for native species. The degree of preference for native species was related to the length of residence in Tucson. Zube et al. (1986) note that Tucson preceded Phoenix in their acceptance of native plants. The Sonoran desert is a unique ecosystem and new residents are often struck by the lack of vegetation. However, this ecosystem has a complex beauty and like an acquired taste may require time and exposure. Therefore, we hypothesize that homeowners who have lived in the Phoenix area for a longer period of time will be more likely to prefer desert dreamscapes than more recent arrivals.

We also predict that homeowners will prefer desert dreamscapes if they have a greater concern about the area's future water supply. Kosek (2003) discovered that residents with memberships in environmental organizations were more likely to prefer significant areas of native vegetation in a residential landscape. While research (Martin, 2001) has challenged the belief that desert landscaping actually reduces water consumption compared with non-native lawn and exotic plants (due to human application error), we hypothesize that homeowners who are more concerned with future security of the Phoenix water supply will be more likely to prefer desert landscapes.

Finally, evidence also suggests that landscape preferences may vary in relation to activities (Zube et al., 1983, Purcell et al., 1994). Purcell et al. (1994) asked respondents to state their preferences for the same set of landscape images based on their (1) general preference, (2) preference for the scene as a place to live or work and (3) preference as a place to visit on vacation. The resulting degree of preference varied according to the qualifying use and suggests ‘cognitive coding’ influences landscape preference. A number of Phoenix residents seek out the desert preserves for regular recreation. Therefore, in addition to visibility (front versus backyard), class, length of residency and water concern, we anticipate that individuals who frequently pursue outdoor recreational activities in natural desert areas will be more likely to prefer desert residential landscapes.

In summary, we hypothesize that landscape preference will vary according to front and backyard location and by income. Higher-income homeowners are more likely to change their residential landscapes to reflect changing fashions and may serve as cultural sentinels for future landscape trends. We predict that homeowners who have lived in Phoenix for a longer period of time will be more likely to prefer desert landscaping than more recent arrivals. We anticipate that homeowners who are more concerned about the security of the future water supply will be more likely to prefer desert landscapes, and finally homeowners who pursue outdoor recreational activities that involve the natural desert environment will be more likely to prefer desert landscaping.

In this study, we are interested in assessing the consistency of landscape preferences with landscape behaviors and investigating whether macro-level forces and micro-level preferences determine landscape behavior. de Groot and van den Born (2003, p. 137) provide a useful metaphor of a ladder to explain how the notion of ‘landscape’ transitions from the imagined to the preferred to the realized. de Groot and van den Born's landscape preference ladder has three rungs. The first rung of the ladder investigates landscape preference in a very general or abstract sense using verbal descriptors, such as ‘untamed nature’. The second rung of the ladder asks subjects to evaluate landscape images for preference. The third rung of the ladder examines “preferences for landscapes as expressed in concrete behaviors of daily life”. This research focuses on the transition from the second rung to the third rung and attempts to untangle the individual level characteristics from the larger social contextual characteristics that shape residential landscape behavior.

The relative importance of the residential landscape needs to be placed within the context of buying a home. Many prospective buyers may view the front and backyard landscapes as more flexible and less expensive to change compared to structural components of the house. Anyone purchasing a home realizes this decision involves a series of trade-offs. While price is often the greatest constraint, other trade-offs include house size and condition, interior and exterior appearance, location, quality of the school district, investment potential, perceived safety, neighborhood image, and proximity to recreational amenities, shopping facilities and the workplace. As part of a study designed to assess whether developers were providing buyers with their desired array of house styles, 129 middle-income homebuyers in Buffalo, New York were asked to prioritize their trade-offs. “Overall, style ranked ninth in importance. The most important factor was price, followed by house quality, the area, maintenance and durability, resale and investment value, site, neighborhood amenities and privacy” (Langdon, 1982, p. C12). Langdon's study, which also concluded that builders were providing a viable range of house styles from which to choose, indicated that the appearance of the residential front and backyard landscapes are less important than the characteristics of the house.

The home building industry is the largest economic driver in the Phoenix economy (Luckingham, 1989, Gammage, 1999, Kirby, 2000, Kirby et al., 2003). As throughout the United States, single developers often design and construct large residential subdivisions. In Phoenix, however, the availability of land has permitted extremely large-scale developments. One development under construction (as of 2005), Vistancia, located at the northwest edge of the metropolitan fringe, will consume 7100 acres and accommodate 45,000 housing units upon completion. When buying a newly constructed home, the buyer typically selects the house model, the lot, and predetermined elements. These predetermined elements include a subset of choices for floor coverings, appliances, cabinets, counters, as well elements for the front and back landscape. While the back landscape may be optional, most developers require the completion of the front yard landscape and carefully prescribe a very limited number of landscape options for the buyer to select from. While ‘up-grades’ (part of the vernacular language for new home buyers) may be possible, the overall front yard dreamscape style will be set by the developer in an effort to provide visual consistency throughout the neighborhood.

So in an effort to sell houses, developers anticipate homeowners’ tastes and in effect, package their homes within desirable dreamscapes. These anticipated preferences change over time in an effort to visually distinguish new homes from nearly new homes and encourage consumption. In the 1930's–1940's, the oasis landscape was extremely popular. In homes dating from the 1950's to 1960's, the lawn landscape dominated. More recently, the desert landscape and courtyard landscape typologies have gained popularity in new housing developments. Each landscape typology can be an example of a dreamscape and we hypothesize that the legacy of the developer's dreamscape, as indicated by the year of construction, will have a greater impact on landscape behavior in the front yard than it will in the backyard.

Section snippets

Study area

Frey (2002, p. 350) has called Arizona one of the New Sunbelt states. He believes that settlements in the New Sunbelt “might be characterized as America's suburbs because of the demographic dynamics that are creating its growth… [and] over the 1990's, domestic migrants to the New Sunbelt outpaced immigrants by a ratio of five to one”. The majority of new arrivals in the Phoenix area (which contains approximately 70% of the state's population) are young married couples with children seeking good

Results

As hypothesized, landscape preference varied significantly according to front and backyard location (Table 2). In the front yard, the column percentages show that 38% of the 230 respondents preferred desert landscaping, followed by 32% of respondents who preferred oasis landscaping. In the backyard, the row percentages show that oasis landscaping was preferred by 39% of all respondents. Only 13% of the respondents preferred desert landscaping in the backyard. Fifty-five percent of the

Conclusions and discussion

Landscape preferences and landscape behaviors for front and backyard residential landscapes vary in a manner that is consistent with Goffman's ideas about the symbolic presentation of self. In the visible front yard, desert landscaping (perceived by most as more socially correct) was the more frequent preference, while in the less visible backyard, the luscious, water-consumptive oasis landscape was much more highly favored. Therefore, we propose that in the front yard, form follows fashion

Acknowledgements

This Project was supported by grants from the National Science Foundation's (NSF) Biocomplexity in the Environment program (SES 0216281) and the Central Arizona-Phoenix Long-Term Ecological Research project (DEB 97114833). SLH was also supported by a sabbatical fellowship at the National Center for Ecological Analysis and Synthesis, a Center funded by NSF (DEB 9421535), the University of California at Santa Barbara, and the State of California. Any opinions, findings, and conclusions or

References (34)

  • D. Brown

    Thorstein Veblen in the Twenty-First Century

    (1998)
  • M.J. Dear

    The Postmodern Urban Condition

    (2000)
  • J.S. Duncan

    Landscape taste as a symbol of group identity: a westchester county village

    Geograph. Rev.

    (1973)
  • W. Frey

    Three Americas: the rising significance of regions

    J. Am. Plan. Assoc.

    (2002)
  • G. Gammage

    Phoenix in Perspective: Reflections on Developing the Desert

    (1999)
  • E. Goffman

    The Presentation of Self in Everyday Life

    (1959)
  • Harlan, S., Rex, T., Larsen, L., Hackett, E., Kirby, A., Wolf, S., Bolin, B., Nelson, A., Hope, D., 2003. The Phoenix...
  • Cited by (196)

    View all citing articles on Scopus
    1

    Tel.: +1 480 965 6213(O)/756 6646 (R); fax: +1 480 965 7671.

    View full text