Cogn Process DOI 10.1007/s10339-014-0607-3
RESEARCH REPORT
Acquisition of spatial knowledge in different urban areas: evidence from a survey analysis of adolescents Ģirts Burgmanis · Zaiga Krišjāne · Jurģis Šķilters
Received: 31 March 2013 / Accepted: 10 March 2014 © Marta Olivetti Belardinelli and Springer-Verlag Berlin Heidelberg 2014
Abstract We herein explore the perception of the geographic environment and analyse the mechanisms that constrain the cognitive processing of spatial information in general. Our guiding theoretical background assumption is that the structure of the spatial environment is a cognitively robust and mutually constrained threefold system relating (1) cognitive topology (comprised of a path and place structure of spatial information and constrained by reference frame-based factors), (2) experience-based functional knowledge (including the effects of socio-economic factors, frequency and familiarity) and (3) linguistic representations (primarily encoded in the prepositional system of a natural language). Here, we focus on (2), i.e. the effect of functional knowledge on the process of acquiring spatial knowledge. We empirically tested adolescents aged 12–17 years to explore the interaction between frequency, familiarity and functional knowledge from a developmental point of view. The social factors we explore are precisely defined and parameterized in our results (exposure to a particular urban area, place of residence, gender, age and factors relating to the environmental and social quality of the local area). Our research shows that there are divergences between the socalled objective topology and the cognitive typology of the urban environment that are significantly constrained by intensity of interactions with environment, number of functionally significant places within particular area and
Ģ. Burgmanis (&) · Z. Krišjāne Faculty of Geography and Earth Sciences, University of Latvia, Alberta Street 10, Riga 1010, Latvia e-mail: [email protected] J. Sˇk¸ilters Center for Cognitive Sciences and Semantics, University of Latvia, Riga, Latvia
age from a developmental perspective in terms of spatial knowledge acquisition. Keywords Spatial knowledge · Spatial cognition · Urban environment · Anchor points · Social value mapping
Introduction The processing of spatial information or, using the term of Downs and Stea (1973), ‘cognitive mapping’ through perception develops dynamic mental representations and transformations of the real and physical world, enabling humans mentally to structure a variety of place types and space in general. Although cognitive mapping allows an individual to build a coherent cognitive structure of an environment, the formation of this structure is constrained by three factors that interact with each other and are inherent in the spatial cognition: the frame of reference in which that individual is located (Coventry and Garrod 2004; Herskovits 1986; Landau and Jackendoff 1993; Zwarts 2005), the functional knowledge and individual acquires and uses in every day spatial processes (Coventry and Garrod 2004) and the language that is used in communicating their spatial experience (Coventry and Garrod 2004; Herskovits 1986). In the present study, we focus on the second system mentioned above, that of functional knowledge assuming that such spatial constraint is among the factors that determine the transformation of the physical space into the cognitive. We elaborate the variety of functional knowledge effects that determine spatial cognition in adolescents. In particular, we focus on the effects of frequency and familiarity as well as urban form on the acquisition of spatial knowledge within urban environment.
123
Cogn Process
Acquisition of spatial knowledge Studies on human spatial cognition mostly support the idea that the extent of spatial knowledge increases with the increase in familiarity with a particular area (Golledge 1978; Golledge and Spector 1978; Montello 1998; Siegel and White 1975). In other words, humans accumulate spatial knowledge through their spatial experience of acting in and interacting with the environment. This in turn means that familiarity determines not only the quantity, but also the quality of spatial knowledge about locations, distances and directions on various scales. Recent studies show that the spatial cognition is shaped by interactions of spatial and social knowledge (Maddox et al. 2008; Taylor et al. 2011). Therefore, the quality of spatial knowledge determined also by acquired information on social attributes of space including values, meanings, social interactions and peoples. The classical theoretical fundamentals of acquisition of spatial knowledge in large-scale environments are developed by Siegel and White (1975). These authors propose that internal representations of spatial knowledge of a place develop in three hierarchical stages from landmark recognition and development of knowledge about paths between landmarks to survey knowledge. In contrast to the theoretical framework put forward by Siegel and White, Montello’s (1998) ‘continuous framework of spatial microgenesis’ focuses on cognitive mapping as a continuous process and draws attention to the importance of familiarity and to individual differences in the acquisition of spatial knowledge and the generation of spatial representations. Spatial knowledge accumulation or increase in familiarity has progressive nature which is instantiated in the quantity, accuracy and completeness of spatial knowledge as an individual’s link with a place becomes stronger. Montello (1998) also proposed that individual differences may exist in the acquisition of spatial knowledge and that socio-demographic characteristics of individuals can determine the process of cognitive mapping.
Cognition of the urban environment The first noteworthy contribution to research on urban cognition was made by architects and urban planners who attempted to improve the visual attraction and functionality of the built environment (Appleyard 1973; Lynch 1960). Hence, the exploration of learning experience of urban environment may contribute and improve the basis of knowledge in city planning grounded by prominent urban planners such as Kevin Lynch. The acquisition of spatial knowledge in urban areas depends on an individual’s motion through the
123
environment and is based not only on direct perception, but more often relies on a learning experience (Lynch 1960). Therefore, it is clear that familiarity, which plays a crucial role in the cognition of an urban environment, determines that those objects and elements of an environment that occur more frequently will be more accessible in the human reference system and thus more easily identified later on. Golledge (1978) and Golledge and Spector (1978) drew on this idea of an important connection between motion, acts of navigation and familiarity with a particular urban area to propose the ‘anchor point’ theory. In this theoretical framework, the principal focus is assigned to the relative significance of landmarks, locations, paths and areas in the spatial environment for each individual. Initially, every individual has anchor points or primary nodes such as home, work, school and shops; all of which are important in his or her daily routine. Subsequently, an individual connects those anchor points or places with paths. An essential part of developing spatial knowledge is the search for paths and the journey between anchor points through segments of space. In the process of searching, an individual acquires knowledge and learns about new, oftenvisited places and landmarks defined as secondary, tertiary and lower-order nodes (Golledge and Stimson 1997). As the result of interaction along the paths between the nodes, there occurs a development of areal concepts such as neighbourhood, community and region. Couclelis et al. (1987) extend anchor point theory by suggesting that nodes are not always point-like (landmarks), but may be also be larger-scale areas and paths. Following the ideas proposed by Lynch and by Golledge, numerous studies have confirmed that landmarks as well as other categories of the urban environment (paths, edges, districts and nodes as proposed by Lynch 1960) are significant in learning, structuring and way-finding in urban areas (Aragones and Arredondo 1985; Golledge et al. 1993; Loomis et al. 1999; Wang and Spelke 2000). However, despite the influence and strong tradition of applying Lynch’s categories in urban cognition studies, some recent work has taken a more critical view on this object-centred approach. First, Golledge (1999) reconsiders the urban elements that shape spatial representations of a city and suggests that they should be viewed in a more general way in geometric terms of points, lines, areas and surfaces. Further, several authors emphasize that the urban elements proposed by Lynch are largely oversimplified and instead have to be considered as dynamic and complex entities acquired through interaction with the environment and also includes contextual knowledge consisting of a large amount of information on spatial features, the spatial interaction between users, the spatial conditions of the surrounding area (Stevens 2006; Portugali 2011) and social
Cogn Process
knowledge (Golledge and Stimson 1997; Maddox et al. 2008; Taylor et al. 2011). Therefore, not only single attributes, but also the functionality, symbolic meaning and form of an urban area in which a particular object is located are essential for the development of urban cognition. This approach has been employed and partially confirmed by empirical studies that use virtually generated environments to facilitate the exploration of spatial behaviour in large-scale areas (Newman et al. 2007; Waller et al. 2003). These studies show that the acquisition of spatial knowledge in cities may depend on the urban form. These studies also indicate that knowledge about an urban area can be acquired as a spatial layout as a whole and that small changes to this layout may affect cognitive mapping. In this case, general visual information about and configuration of the urban area play a more significant role than the single properties of the area in the development of awareness about that area. Similarly, Montello and Sas (2006) suggest that urban form affects the acquisition of spatial knowledge, and they conclude that a more articulated space that is broken up into more different parts, such as a built environment with high structural diversity in terms of colour, design and complex layout, is more difficult to perceive and use for routine spatial tasks. The effect of the spatial configuration of area on spatial urban cognition partially explains the ‘space syntax’ theory of Hillier (2012), which posits that there exist general spatial laws in terms of a human’s geometric intuition that are encoded or embedded through planning in the design of city and that at the same time allow an individual to conceive attributes of the urban environment. The line is the most appropriate element for human geometric intuition in that it preserves informational stability for moving individuals as well as visibility and an almost complete picture of an overall system (Hillier 2002). For these reasons, the grid system of street networks (i.e. containing areas with a clear geometry) allows patterns of urban form to be more easily and intuitively perceived and used for movement, as opposed to irregular, asymmetric and more complex urban layouts. This assumption also supported by several empirical works (Wang and Spelke 2002; Yaski et al. 2011).
Factors influencing children’s spatial cognition of the geographic environment Extensive research on children’s cognition and spatial behaviour started in the 1960s with a significant body of work by Jean Piaget (see for example Piaget and Inhelder 1967) that influenced the emergence of cognitive developmental studies in the field of human geography exploring the spatial abilities of young children aged 3–6 years old
(Blaut and Stea 1971; Blaut et al. 2003) and assessing their development of spatial knowledge in a large-scale environment (Hazen 1982; Hazen et al. 1978). Other studies have highlighted the acquisition of spatial knowledge by children aged 5–13 years from many different points of view (Golledge et al. 1985; Matthews 1987, 1992; Rissotto and Tonucci 2002, 2006). Matthews (1992) summarizes previous empirical evidence and indicates that three main factors affect children’s environmental cognition: active and passive exploration of the environment, the nature of activity and the level of activity control. More active and less controlled exploration of the accessible environment leads to more spatial representations. Matthews (1987) indicates that age and gender have an effect on the level of children’s spatial experience. By assessing free recall sketching, map interpretation and aerial photograph interpretation he found that the size of the ‘free home range’ of children aged 6–11 years is directly linked with their knowledge about the environment. Children aged 6–9 years and girls, who have stronger spatial restrictions, showed weaker awareness of the environment near to their place of residence. Similar results were found by Golledge et al. (1985), who used the route navigation of children aged 9–11 between two points as the basis for their research. The authors observed that children who are residents of the same neighbourhood have different levels of spatial knowledge that depend on the number of major nodes in the children’s route. Younger children with a limited free range in their residential neighbourhood cannot identify as many locations as those who have a larger daily living space. Also, children recognize a greater number of familiar locations more quickly and more accurately if they are experienced daily. Several empirical studies have extended the research on gender differences in the mental representations of space by focusing on the effect of children’s mode of travel to activity sites and the level of independent mobility on the development of spatial cognition. Garcia-Mira and Goluboff (2005) reported that there are different ways of conceptualizing the same space that depend not only on the age and gender of the children, but also on the mode of transport used. Children who walk to school acquire more information about their environment, remember more places and objects and demonstrate a greater ability to structure space. Rissotto and Tonucci (2002) explored independent/dependent mobility and confirm the importance of an active individual—the level of interaction with the environment acts as a constraining factor in children’s acquisition of environmental knowledge. They found differences in the extent of spatial knowledge that can be associated with the mode of children’s travel behaviour. Children who travel to school independently had better
123
Cogn Process
scores in the tasks used in the study (drawing a sketch map, reading a blank map and drawing a home to school itinerary map). Despite the long-standing tradition in cognitive geography and developmental psychology of focusing on exploring how young children acquire spatial knowledge, in our study we examine children aged 12–17 years. There are four key reasons for our choice of this particular age group. First, although children’s understanding and representations of space as well as their behaviour in space are of great interest to geographers, psychologists, cognitive scientists and environmental planners, there are not many studies that consider children’s acquisition of spatial knowledge in large-scale environments between the ages of 12 and 17. Second, from the development perspective, this age group corresponds to a transitional stage of human development— adolescence—when substantial socio-psychological changes (e.g. in the development of identity) occur, as do major changes in spatial behaviour through an increase in spatial autonomy. For instance, the spatial restrictions established by parents for growing adolescents gradually decrease, and the mode of mobility becomes more independent (O’Brien et al. 2000). Hence, this is the period in human life when a substantial transition from a conceptual to more functional knowledge of large-scale space occurs. Third, following Sibley (1995), we agree with the proposition that adolescents are the most spatially marginalized group in the modern urban environment. Hence, we would argue that it is crucial that we explore how the various modes and intensity of environmental interaction, as well as the daily experienced and accessible functionally and physically diverse urban environments, affect an adolescent’s spatial cognition of the geographic environment, in order to show the influence of modern urban life on adolescents’ cognitive development in general terms. Finally, we consider that the age group of 12–17-yearold adolescents may be a very useful and indeed more valuable source of data than older age groups for explaining divergences between the so-called objective topology of the environment and the cognitive topology of the environment, as well as eliciting the factors that may constrain the accurate perception of a real-world environment. This assumption is based on the thesis that children and adolescents mostly have not experienced significant transformations of the environment (e.g. change of place of residence, less travel experience).
Social value mapping technique Instead of traditional tools used for exploring external representations of stored spatial knowledge such as sketch
123
maps, map interpretation and verbal descriptions, we used the social value mapping technique (Tyrvainen et al. 2007) as our main tool for measuring the extent of respondents’ spatial knowledge. We consider the social value mapping task to be an appropriate method for exploring adolescents’ knowledge about space and their perception of the environment as well as the way in which the perceived real world is transformed into a cognitive format and the factors that influence this transformation. This opinion is based on two assumptions. First, as it was mention above, the social knowledge of space including attributive values and attitudes interacts with spatial (geometrical and topological) knowledge and, thus, together generate spatial cognition (Golledge and Stimson 1997; Maddox et al. 2008; Taylor et al. 2011). The second is grounded in the conceptual understanding of ‘place’ in human geography where it is defined as a part of space with human-assigned meaning (see, e.g. Agnew 1987; Cresswell 2004; Tuan 1975, 1977). Consequently, humans assign meanings primarily to functionally relevant parts of experienced space (for a slightly different point of view cf. Coventry et al. 2010; Coventry and Garrod 2004) in large-scale environments. Furthermore, functionality of place in an objective geographic environment can be mapped cognitively faster than other information, Based on this notion and developmental characteristics of age groups, we assume that social values such as safe/unsafe, pleasant/unpleasant and freedom from adult supervision clearly representing functionality of place can all be important for adolescents when they choose places in an urban environment to go about their daily social lives. It is therefore logical that those parts of the urban environment with such additional and relevant information can be identified by adolescents as a ‘places’ and can be cognitively mapped faster than other places in a particular area. These crucial aspects of human cognition determine the strength of social value mapping techniques that allow the storage of spatial experience, knowledge of the respondent and the building of internal representations of the external environment. When respondents were asked to assign numbers for social values to places on a real map they had to (a) use their stored spatial knowledge about the local environment, (b) search mentally for the places in their cognitive map where functionality corresponds to the conceptual meanings of social values and (c) build an internal cognitive representation of the environment. The present study focuses on the exploration of functional knowledge factors that may constrain the acquisition of spatial knowledge within a city, an environment with high and continuously growing complexity in terms of its physical and socio-economic structure. Having considered previous studies (see, e.g. Couclelis et al. 1987; Matthews 1987, 1992; Montello and Sas 2006;
Cogn Process
Yaski et al. 2011) and fundamental theoretical frameworks (Golledge 1978; Montello 1998; Siegel and White 1975), we propose following hypothesis. First, we assume that socio-demographic differences, duration of living in a specific place of residence, frequency of travel and number of activities in a particular neighbourhood are significant factors that can affect an individual’s ability to acquire a diverse amount of spatial knowledge concerning locations in the vicinity of residence. Second, we also believe that, in addition to familiarity with a particular area, which is gradually formed according to the amount of interaction with the accessible environment, humans acquire spatial knowledge about an urban area as a spatial layout as a whole and the differences in the urban form may have an effect on the generation of spatial representations. Therefore, this study tests a variety of variables: (1) time of residence in a particular neighbourhood/urban area, (2) urban form of the area, (3) frequency and type of environment usage, (4) mode of travel between activity places and (5) age and gender with familiarity of environment.
Case study areas The present research was undertaken in three study areas located in different parts of the largest city in Latvia, Riga. The areas were chosen so that each represented specific type of urban environment was physically and functionally distinct. Each of these areas contains two or more neighbourhoods that have similar physical settings. The first area (Area 1) is located within the central part of the city and can be characterized as an extension of the city centre (which is informally called ‘the silent centre’). The territory of Area 1 occupies 6.5 km2 and contains around 15,000 inhabitants in total (Riga City Council 2008). The area has a medium-density built environment where residential houses are mixed with diverse commercial, industrial and public service buildings. Most of the houses were built at the beginning of the twentieth century and are five-storey buildings. However, in the 1960s, a number of new residential multi-storey houses were built, thus bringing in an additional number of residents from a different socio-economic and socio-demographic background. The monocentric spatial and functional structure of Riga is grounded in its historical growth, which means that Area 1 contains more functionally significant objects and places than the two other areas covered by this study. Area 1 also has a more dense, irregular street network. The second area (Area 2) is located on the outer edge of the city on the right bank of the river Daugava and comprises three neighbourhoods—Purvciems, Plavnieki and Dreilini. The urban environment of the neighbourhoods included in Area 2 is similar to the residential areas of
other post-soviet cities where wide expanses of large-scale housing estates rarely overlap with small areas of green space and mixed-use areas. The construction of buildings mainly took place between the 1970s and the 1990s, a period corresponding to Soviet-era urban planning, architectural and construction outlines and typified by a high density of 5–16-storey reinforced concrete apartment houses that make the physical and functional structure of the area markedly oppressive and homogenous. Area 2 is therefore also characterized by a shortage of prominent objects and places of cultural and historical significance as well as a less-dense and more regular network of major streets. Area 2 is enclosed by major roads, which are also seen as barriers that prevent people from freely reaching and using the local public and green spaces, thereby restricting the spatial experiences of younger people due to safety concerns. Area 2 has about 85,000 inhabitants and covers an area of 13 km2 (Riga City Council 2008). The third area (Area 3) is located in the western part of Riga on the left bank of the river Daugava, and in this study represents a mixed built and mixed land-use urban environment. Area 3 consists of three neighbourhoods— Zolitude, Pleskodale and Sampeteris—with a total area of about 12 km2, all significantly different in terms of urban form. This means that these three parts of Area 3 have different urban density and make-up. Area 3 mostly consists of an industrial zone and a small residential zone where wooden one-storey houses from the beginning of the twentieth century coexist with houses that are no higher than three storeys, which were built between the 1950s and the 1990s. The more homogeneous part of Area 3 is situated in the south and is dominated by detached family houses. A remarkable feature of the southern part of Area 3 is that between the clusters of one-to-three-storey detached houses built during the twentieth century, there are also green areas of various sizes. A typical residential zone of Riga is located in the northern part of Area 3. It is a highdensity territory with 6–16-storey apartment houses built in the 1980s. In this part, the commercial and public service buildings are distributed along the large-scale housing estates and close to the major roads. Approximately 19,000 out of the 30,000 inhabitants in Area 3 live in these largescale housing estates (Riga City Council 2008) in the northern part of this study area. The choice of Areas 2 and 3 was based on the fact that, in contrast to Area 1, the locations of these two areas largely restrict opportunities for residents to reach essential objects with diverse functionality and the city centre with ease. Our selection of three such diverse areas was based on the assumption that the different locations of the respondents’ places of residence in the various urban environments and their access to essential objects and the city centre may affect their generation of representations of space.
123
Cogn Process Table 1 Socio-demographic characteristics of sample (N = 449) Characteristics
Area 1
Area 2
Area 3
Age (SD)
14.4 (1.6)
14.1 (1.6)
14.6 (1.4)
Female
54
48
42
Male
46
52
58
Primary
83
84
81
Secondary
17
16
19
96
97
73
4
3
27
N = 111
N = 184
N = 154
Sex (%)
Educational level (%)
Housing type (%) Apartment Private house Total sample
Methods Participants In February 2010, questionnaires were distributed to 942 students aged 12–17 years in three randomly sampled urban schools in the three areas described above. The socio-demographic characteristics of the sample are given in Table 1. Students who completed the questionnaire but stated that they lived outside the study areas were eliminated from the analysis, which meant that 468 (50 %) of the questionnaires were initially included; however, 449 (96 %) of these were accurately completed by the respondents and considered valid. Materials and procedure Participants completed a questionnaire-based survey consisting of three parts from which general and site-specific information about their spatial experience was gathered. Previous studies have considered the role of individual differences in the acquisition of spatial knowledge. We therefore included questions asking respondents for general socio-demographic information (age, sex, education level, housing type). The second part of the questionnaire was designed to elicit information about the respondents’ spatial behaviour in their place of residence (mode of travel to school, length of living in place of residence, amount of leisure time spent within the area daily). Most of the acquired information showed the respondent’s exposure to a particular urban area. This set of questions was designed to test the role of familiarity in the respondents’ cognition of geographic environment as discussed in several classical studies (Montello 1998; Siegel and White 1975). In the third part of questionnaire, we included two questions that would enable us to examine the anchor point hypothesis of Golledge (1978) and the importance of functionally
123
significant places in urban spatial cognition. We asked respondents to indicate on a map of the area the locations of (1) the place of residence of their best friend and (2) their main place of out-of-school activity. If these places were outside the area shown on the map or did not exist, the respondent ticked the box ‘not in this area’. These questions were included because we suppose that, besides home and school, these two places are the most important anchor points in the life of an adolescent and thus including them in our analysis would allow us to examine the effect of functionally significant places in the acquisition of spatial knowledge of this specific age group. The fourth part of the survey included questions aimed at capturing adolescents’ spatial knowledge of the area surrounding their place of residence by eliciting specific information about the particular social values attached to personally significant places. For our social value mapping task, which was originally a GIS-based method developed for spatial planning, we used 1:13,500 scale maps of the areas around the schools involved in the survey. The social values of places were created as conceptual constructs for the survey. This meant that the researchers had to formulate concepts of the environmental qualities of places that were then included in the questionnaires given to the respondents (Makinen and Tyrvainen 2008; Tyrvainen et al. 2007). Unlike the questionnaire in Tyrvainen et al. (2007), where the respondents were asked to assign an ordinal number of an area from the map of the case study area to a particular social value (e.g. a lovely view), we took a slightly different approach in that our respondents were asked to assign independently an ordinal number of social value to those places that in their opinion corresponded to the given value descriptions. Specifically, the respondents were asked to indicate on the map of their area of residence those places they considered pleasant, unpleasant, safe, unsafe and where they felt free from adult supervision (assigned ordinal numbers of social value to particular places as 1–5). There was also a limitation—only one place could be given one of the above five values. If the place the respondent had in mind was not on the map or they were unsure that such a place existed, they ticked the box ‘not in this area’ or ‘hard to say’. Data analysis In our analysis, we coded eight independent variables (sex, age, length of living in place of residence, travel mode to school, best friend’s place of residence within area, main place of out-of-school activity within an area, child’s level of independent mobility and amount of leisure time spent within the area daily), all of which we assume hypothetically may affect a human’s perception of their local urban environment. Statistical analysis of data was performed for
Cogn Process Table 2 Respondents’ spatial knowledge of the urban area surrounding their place of residence and school Independent variables
Area 1
Area 2 H
df
M
Area 3
M
SD
Girl
.54
.37
Boy
.54
.35
12–13
.44
.36
.47
.35
.42
.39
14–15
.54
.35
.41
.39
.50
.36
16–17 .68 .34 Duration of living in place of residence
.41
.38
.58
.37
.42
.38
.53
.39
.43
.37
.51
.36
SD
H
df
M
SD
H
df
.47
.39
1.188
1
.54
.35 5.49
2
.07
1
1.45
1
8.72**
1
.66
1
.516
2
Sex .01
1
.41
.40
.44
.34
.75
1
Age (years)
Less than 5 years
.39
.32
More than 5 years
.63
.35
Active
.56
.36
Passive
.52
.36
7.23*
11.08**
2
1
.41
.02
2
1
Travel mode to school .31
1
.43
.40
.43
.31
.10
1
.49
.38
.57
.34
Best friend’s place of residence within area Yes
.61
.34
No
.47
.36
4.08*
1
.49
.36
.29
.37
12.04**
1
.57
.36
.37
.35
Place of out-of-school formal activity within area Yes
.55
.36
No
.54
.36
.06
1
.32
.31
.45
.38
2.20
1
.55
.38
.50
.36
Child’s level of independent mobility (travelling alone to school) Low
.58
.39
Average High
.45 .55
.29 .36
1.13
2
.46
.36
.49 .42
.31 .38
1.38
2
.53
.33
.56 .50
.38 .38
Time spent on leisure activities within area Less than 2 h
.46
.37
.35
.35
.42
.39
2–8 h
.50
.37
6.82*
2
.45
.38
1.84
2
.50
.38
More than 8 h
.69
.30
.47
.37
.58
.37
2.98
2
Mean scores of the spatial knowledge index. Scores range from 0 to 1 Statistically significant at ** p \ .01; * p \ .05
each study area separately. The main aim of such an approach was to assess the potential effect of the area’s environment including location and physical and social setting as well as other factors that may have an influence on a human’s acquisition of spatial information. To study the factors influencing environmental cognition, we designed an additive index of spatial knowledge that included five variables obtained from the social value mapping task, namely pleasant (1), unpleasant (2), safe (3), unsafe (4) and areas where respondents feel free from adult supervision (5). The questionnaire answers were given 1 point if a value number (1–5) was assigned to place marked on the map. If a social value number was not linked to a place on the map and the box ‘not in this area’ or ‘hard to say’ was ticked, the answer was given a score of 0. Each respondent could thus accumulate a maximum score of five points for this aspect of the questionnaire. Next, for each
individual, we calculated the additive index of spatial knowledge (mean value total points). Thus, the level of each individual’s spatial knowledge could range between 0 and 1. The acquired index was used as the dependent variable in further analysis. The purpose of constructing a spatial knowledge index was to determine whether or not individual differences were shaped by adolescents’ sociodemographic background, their distinct spatial behaviour, or their functional interaction with the urban area have an affect on their amount of spatial knowledge.
Results and discussion The influence of background characteristics as independent variable-based differences in the mean values of the additive index of spatial knowledge was confirmed by the
123
Cogn Process
Kruskal–Wallis test (see Table 2). Table 2 illustrates the differences in the extent of the adolescents’ knowledge about the area surrounding their place of residence and school only. Foremost, we tested the hypothesis stating that (a) sociodemographic differences, (b) duration of living in a specific place of residence, (c) frequency of travel and number of activities in a particular neighbourhood, (d) factors representing familiarity have effect on individual’s ability to acquire a spatial knowledge concerning locations in the vicinity of residence. From Table 2, it can be seen that not all the independent variables that are statistically significant in terms of differences in the mean values of the additive index of spatial knowledge are appropriate for all studied areas. In Area 1, four factors (age, duration of living in place of residence, best friend’s place of residence within area and amount of leisure time spent on leisure activities within area) show stage progressing changes of spatial knowledge. In other two areas characterized as urban environments with a homogeneous physical spatial setting, only the presence of best friend’s place of residence within area influences adolescent’s extent of spatial knowledge. Although age as a factor affecting knowledge about space has usually been considered in studies on young children (Matthews 1987), we show that age also plays a role in the case of adolescents. We find significantly progressive differences in the number of places indicated by respondents, which is reflected in fluctuations in the additive index of spatial knowledge. We clearly distinguish differences between three age groups (12–13 year olds, 14– 15 year olds and 16–17 year olds) and therefore conclude that, like younger children, adolescents also acquire more knowledge about space with age. However, we would argue that the changes in the additive index of spatial knowledge can be attributed not only to increasing knowledge about space as such, but also to the developmental peculiarities of each individual and the gradual change in the way in which adults perceive the surrounding environment. An important factor that affected the adolescents’ spatial experience and increasing familiarity with their local urban environment was the duration of living in the place of residence. Respondents who had lived longer in the neighbourhood could identify a significantly greater number of the places mentioned in the questionnaire. This indicates that the cognitive map of the neighbourhood for this group is developed in more detail than for those who had moved to the area more recently. The results show that it is possible to identify an approximate time boundary beyond which the spatial experience of the individual becomes sufficient for fixing a cognitive topology and representing it in the perception of certain places within the area of residence. The results also show that, in our study,
123
changes of perception due to the accumulation of spatial knowledge are observable after 5 years of living in a particular place of residence as supposedly the territory becomes more familiar. Another important factor that reflects familiarity with the local environment is the amount of leisure time an individual spends in the area of residence outside school and home. Our results show that the more time an individual spends outside home and school, the better their knowledge about the respective territory, thus affecting the quantity and quality of the experience that the individual has of the surrounding territory. Those respondents who spent much of their time exploring the surrounding territory were better able to identify places and rate them according to social values. The location of the place of residence of the respondent’s best friend in the same area as the respondent was included as an independent variable in our analysis based on two considerations. First, we assume that the friend’s place of residence acts as another functionally significant and meaningful place—as an anchor point or primary node —in the adolescent’s objective topology of the local environment and also in the cognitive topology of that environment. The second consideration emerges from the first, namely that the presence of an additional meaningful place or node in the local environment is expected to increase the respondent’s familiarity with that environment. We offer the following explanations for this assumption. The increase in familiarity is associated with the habit of adolescents of going home from school in the company of a friend or friends. This habit can provide the adolescent with an opportunity regularly to visit additional places in the local area, varying the route based on their friends’ important places. Also, the relatively short distance between the respondent’s place of residence and that of their best friend (mostly within 1 km) allows them to reach each other easily and may be regarded as another explanation for the growing adolescent’s greater familiarity with their area. More frequent meetings provide adolescents with more opportunities to spend leisure time together and explore and experience the local environment, which facilitates the acquisition of more detailed spatial knowledge. Evidence to support our assumption is provided by the significant differences in the mean values of the additive index of spatial knowledge between respondents whose friends live in the same area and respondents whose friends do not. It is clear that the presence of the best friend’s place of residence in the same area is a very influential factor in an adolescent’s development of a more accurate cognitive topology and also applies in the case of areas with a spatially homogeneous physical setting (Areas 2 and 3).
Cogn Process Table 3 Relationship between spatial knowledge index and background variables: Spearman rank-order correlation coefficient (r) Independent variables
Area 1 rs
Sex
−.01
Area 2 rs
Area 3 rs
.064
.086
Age
.26**
−.083
−.133
Duration of living in place of residence
.32**
.010
−.022
Travel mode to school
.05
−.023
−.096
Best friend’s place of residence within area
.19*
Place of out-of-school formal activity within area Child’s level of independent mobility (travelling alone to school)
.02
−.109
.065
−.01
−.083
−.047
.087
.139
Time spent on leisure activities within area
.24*
.256**
.236**
Statistically significant at ** p \ .01; * p \ .05
Next, Spearman rank-order correlation analysis was used to identify the relationship between the spatial knowledge index and the background/independent variables. The relatively significant correlations presented in Table 3 point to correlations between the above mentioned four factors of age, duration of living in place of residence, best friend’s place of residence within area and amount of leisure time spent within the area daily, and the extent of spatial knowledge the individual has about the areas of residence. Only one of these four factors, namely best friend’s place of residence within area, has a significant effect on the increase in the respondents’ spatial knowledge in Areas 2 and 3. The correlation values in Table 3 show that with the increase in time that individuals had lived in the area and in time they had spent in the vicinity, there was a corresponding increase in their spatial knowledge of the environment. The increase in knowledge can be explained by a progressive increase in familiarity with the space— identification of routes, physical objects and peculiarities of the environment. Conversely, adolescents with limited leisure time spent in the territory of residence or a shorter duration of living in the neighbourhood showed poorer knowledge in the social value mapping task. These two factors must therefore be viewed as important constraints in terms of the adolescent individual’s perception of the environment and spatial representations. The correlations in Table 3 also show that older adolescents (16–17 years old) have a better knowledge of the area and are more fixed in the social values they assign to these places than the younger ones. The findings also show that in this age group, the influence of gender on the amount of spatial awareness is not significant.
Finally, using the previously mentioned correlations, we tested hypothesis stating that (a) humans acquire spatial knowledge about an urban area as a spatial layout as a whole and (b) the differences in the urban form may have an effect on the generation of spatial representations. The presented correlations once again confirm the importance of the presence of an additional functional anchor point or node for the acquisition of spatial knowledge in an urban setting that is homogeneous in terms of the built environment and has few functionally significant objects. This type of urban environment is observable in the majority of Riga’s residential areas and, in the case of this study, is represented by Areas 2 and 3 in Table 3. The results of the correlation clearly show that there exists a relationship between the urban form of the research area and the formation by the respondents of a cognitive topology of the environment of the area in which they live. We consider that this relationship can be explained by the spatial functional structure of the city of Riga, which is monocentric in form. The distance from the place of residence to the urban core plays an important role in the acquisition of spatial knowledge; the closer the adolescent’s place of residence is to the city centre, the more functionally significant objects are found in the immediate area and located close to each other, and the more dense and irregular the major street networks are. Consequently, due to this spatial pattern of objects, the closer to the city centre the individual lives, the more time they need to acquire spatial knowledge of those objects and the more nodes or meaningful places those individuals may have in the vicinity of their place of residence.
General discussion Our study focused on the ways how functional knowledge (Coventry and Garrod 2004) (as linked to frequency, familiarity and urban form) constrains and determines the transformation of physical space into cognitive one. The use of social value mapping techniques (Tyrvainen et al. 2007) has provided evidence in our study on how an individual modifies the physical urban environment into a cognitive map and how this depends on factors such as familiarity with a particular urban environment and age. First, our findings support our assumption that the duration of living in the place of residence and the amount of leisure time spent in the area are directly associated with the amount of acquired spatial knowledge. An increase in an adolescent’s age also leads to an increase in their familiarity with the surrounding environment and overall knowledge of spatial relationships. These findings are particularly consistent with the theoretical approach of
123
Cogn Process
Montello (1998), who emphasizes that the acquisition of spatial knowledge about a particular area is a continuous process in human cognitive development. The continuous framework of spatial microgenesis posited by Montello (1998) and in other previous research (O’Brien et al. 2000) also explains continuous changes in the amount of spatial knowledge in the case of adolescents. The extent of spatial experience and therefore amount of spatial representations also continue to increase after the age of 12 due to gradually decreasing restrictions and an increasing amount of independent mobility. We suppose that this result is directly associated with the significant increase in individual differences and personality features, which develop with age and which are inevitably present in an individual’s perception of their environment. The presence of the place of residence of a best friend is indisputably a very strong factor that affects the amount of spatial knowledge, which in turn indirectly affects familiarity with a particular area. Anchor point theory (Golledge 1978) can provide an explanation for this correlation in our study in that Golledge found that the existence of an additional node in the local environment has a significant effect on the formation of cognitive topology (a finding which is also consistent with Golledge et al. 1985). Second, the results of our analysis of the impact of the spatial structure of different urban areas on the acquisition of spatial knowledge prove our hypothesis that the physical form and structure of the urban environment exert a constraining effect on the acquisition of spatial knowledge and the formation of spatial cognition. This finding is partly consistent with previous evidence that supports the opinion that the whole layout of an urban area may influence the acquisition and use of spatial knowledge (Montello and Sas 2006; Yaski et al. 2011). The present study was performed in various three different parts of a city that has a monocentric functional spatial structure, and our findings demonstrate that, in contrast to areas located within the urban core that have a more heterogeneous physical and functional structure as well as a more dense and irregular street network (Area 1), urban areas without notable landmarks and with a homogeneous physical structure including a sparse and regular major street network (Area 2 and Area 3) do not encourage the progressive development of cognitive maps, and moreover partly limit the individual’s ability to generate spatial representations of their environment. Considering these findings regarding the three case study areas and referring to theoretical works of Lynch (1960), Golledge (1999) and Hillier (2012) that support the notion that urban forms generally are perceived as geometric configurations that are generally shaped by lines or paths that are reflected in the urban environment as a street network, we would argue that an urban area with a dense
123
and irregular street network is harder to understand than an urban area with a sparse, linear and regular layout. Furthermore, we would also suggest that, although a higher density of street network adds more complexity to the layout of an urban area, which makes it more difficult to know fully, at the same time, this inherent complexity allows the individual to generate more spatial representations over time. Hence, the more central the place of residence, the more gradual, continuous and complete is the cognitive mapping. To conclude, our research findings strongly suggest that the mapping of objective topology onto the cognitive is enhanced by familiarity with the environment as well as by familiarity with the functionally significant objects in a particular area. The more actively the environment is processed, the richer the representations generated. This finding supports prior research that suggests that active and independent exploration of large-scale environments is important in the acquisition of spatial knowledge (Matthews 1992). In our study, the longer the adolescents have resided in the area and the more frequently they spend leisure time there, the more places they can associate with the prescribed social values. Despite the difference in the nature of our study area (in our case, large-scale spatial environments), we believe our results support the findings of Coventry and Garrod (2004), which show that the underlying structure of spatial cognition is supported by a wide range of functional knowledge and co-influenced by socio-economic, cultural and other factors such as spatial behaviour and transportation mode that then determines the individual’s perception of geographic environment and acquisition of spatial knowledge. Acknowledgments The research was promoted with support of the European Union Social Fund project ‘Support of Doctoral Studies at the University of Latvia’, Project No. 2009/0138/1DP/1.1.2.1.2/09/ IPIA/VIAA/. Jurģis Šķilters’ work is supported by the Fulbright Scholar Program.
References Agnew J (1987) Place and politics: the geographical mediation of state and society. Allen & Unwin, Boston Appleyard D (1973) Notes on urban perception and knowledge. In: Downs R, Stea D (eds) Image and environment: cognitive mapping and spatial behavior. Aldine, Chicago, pp 109–114 Aragones JI, Arredondo JM (1985) Structure of urban cognitive maps. J Environ Psychol 5:197–212 Blaut JM, Stea D (1971) Studies of geographic learning. Ann Assoc Am Geogr 61:387–393 Blaut JM, Stea D, Spencer C, Blades M (2003) Mapping as a cultural and cognitive universal. Ann Assoc Am Geogr 93:165–185 Couclelis H, Golledge RG, Gale N, Tobler W (1987) Exploring the anchor-point hypothesis of spatial cognition. J Environ Psychol 7:99–122
Cogn Process Coventry KR, Garrod SC (2004) Saying, seeing and acting: the psychological semantics of spatial prepositions. Psychology Press, New York Coventry KR, Lynott D, Cangelosi A, Monrouxe L, Joyce D, Richardson DC (2010) Spatial language, visual attention, and perceptual simulation. Brain Lang 112:202–213 Cresswell T (2004) Place: short introduction. Blackwell, Malden Downs R, Stea D (1973) Image and environment: cognitive mapping and spatial behavior. Aldine, Chicago Garcia-Mira R, Goluboff M (2005) The perception of urban space from two different viewpoints: pedestrians and automobile passengers. In: Garcia-Mira R, Uzzell DL, Real JE, Romay J (eds) Housing, space and quality of life. Antony Rowe Ltd., Chippenham Golledge RG (1978) Learning about urban environments. In: Carlstein T, Parkes D, Thrift N (eds) Timing space and spacing time: making sense of time, vol 1. Edward Arnold, London, pp 76–98 Golledge RG (1999) Human wayWnding and cognitive maps. In: Golledge RG (ed) Wayfinding behavior: cognitive mapping and other spatial processes. The Johns Hopkins University Press, Baltimor, pp 5–45 Golledge RG, Spector AN (1978) Comprehending the urban environment: theory and practice. Geogr Anal 10:403–426 Golledge RG, Stimson RJ (1997) The spatial behaviour: a geographical perspective. The Gilfourd Press, New York Golledge RG, Smith T, Pellegrino JW, Doherty S, Marshall SP (1985) A conceptual model and empirical analysis of children’s acquisition of spatial knowledge. J Environ Psychol 5:125–152 Golledge RG, Ruggles AJ, Pelegrino JW, Gales ND (1993) Integrating route knowledge in an unfamiliar neighborhood along and across route experiments. J Environ Psychol 13:293–307 Hazen NL (1982) Spatial exploration and spatial knowledge: individual and developmental differences in very young children. Child Dev 53:826–833 Hazen NL, Lockman JJ, Pick HL Jr (1978) The development of children’s representations of large-scale environments. Child Dev 49:623–636 Herskovits A (1986) Language and spatial cognition. Cambridge University Press, Cambridge Hillier B (2002) A theory of the city as object: or, how spatial laws mediate the social construction of urban space. Urban Des Int 7:153–179 Hillier B (2012) Studying cities to learn about minds: some possible implications of space syntax for spatial cognition. Environ Plann B 39:12–32 Landau B, Jackendoff R (1993) “What” and “where” in spatial language and spatial cognition. Behav Brain Sci 16:217–265 Loomis JM, Klatzky RL, Golledge RG, Philbeck JW (1999) Human navigation by path integration. In: Golledge RG (ed) Wayfinding behavior: cognitive mapping and other spatial processes. The Johns Hopkins University Press, Baltimor, pp 125–151 Lynch K (1960) The image of the city. MIT Press, Cambridge Maddox KB, Rapp DN, Brion S, Taylor HA (2008) Social influences on spatial memory. Mem Cogn 36:479–494 Makinen K, Tyrvainen L (2008) Teenage experiences of public green spaces in suburban Helsinki. Urban For Urban Green 7:277–289 Matthews MH (1987) Gender, home range and environmental cognition. Trans Inst Br Geogr 12:43–56 Matthews MH (1992) Making sense of place: Children’s understanding of large-scale environments. Barnes & Nobles Books, Savage
Montello DR (1998) A new framework for understanding the acquisition of spatial knowledge in large-scale environments. In: Egenhofer MJ, Golledge RG (eds) Spatial and temporal reasoning in geographic information systems. Oxford University Press, New York, pp 143–145 Montello DR, Sas C (2006) Human factors of watfinding in navigation. In: Karwowski W (ed) International encyclopedia of ergonomics and human factors, 2nd edn. CRC Press, Taylor & Francis, Ltd, London, pp 2003–2008 Newman EL, Caplan JB, Kirschen MP, Korolev IO, Sekuler R, Kahana MJ (2007) Learning your way around town: how virtual taxicab drivers learn to use both layout and landmark information. Cognition 104:231–253 O’Brien M, Jones D, Sloan D, Rustin M (2000) Children’s independent spatial mobility in the urban public realm. Childhood 7:257–277 Piaget J, Inhelder B (1967) The child’s conception of cpace. W. W. Norton, New York Portugali J (2011) Complexity, cognition and the city. Springer, Hedelberg Riga City Council (2008) Statistical materials of Riga City Council City Development Departament. Riga City Council City Development Departament, Riga Rissotto A, Giuliani MV (2006) Learning neighborhood environments: the loss of experience in modern world. In: Spencer C, Blades M (eds) Children and their environments: learning, using and designing spaces. Cambridge University Press, Cambridge, pp 75–90 Rissotto A, Tonucci F (2002) Freedom of movement and environmental knowledge in elementary school children. J Environ Psychol 22:65–77 Sibley D (1995) Geographies of exclusion: society and difference in West. Routledge, London Siegel AW, White SH (1975) The development of spatial representations of largescale environments. In: Reese HW (ed) Advances in child development and behavior, vol 10. Academic Press, New York, pp 9–55 Stevens Q (2006) The shape of urban experience: a reevaluation of Lynch’s five elements. Environ Plann B 33:803–823 Taylor HA, Wang Q, Gagnon SA, Maddox KB, Brunye´ TT (2011) The social connection in mental representations of space: explicit and implicit evidence. In: Egenhofer M et al (eds) COSIT 2011, LNCS 6899. Springer, Heidelberg, pp 231–244 Tuan Y-F (1975) Images and mental maps. Ann Assoc Am Geogr 65:205–213 Tuan Y-F (1977) Space and place: the perspective of experience. University of Minnesota Press, Minneapolis Tyrvainen L, Makinen K, Schipperijn J (2007) Tools for mapping social values of urban woodlands and other green areas. Landsc Urban Plan 79:5–19 Waller D, Loomis JM, Steck SD (2003) Inertial cues do not enhance knowledge of environmental layout. Psychon B Rev 10:987–993 Wang RF, Spelke ES (2000) Updating egocentric representations in human navigation. Cognition 77:215–250 Wang RF, Spelke ES (2002) Human spatial representation: insights from animals. Trends Cogn Sci 6:376–382 Yaski O, Portugali J, Eilam D (2011) City rats: insight from rat spatial behavior into human cognitionin urban environments. Anim Cogn 14:655–663 Zwarts J (2005) Prepositional aspect and the algebra of paths. Linguist Philos 28:739–779
123
RESEARCH REPORT
Acquisition of spatial knowledge in different urban areas: evidence from a survey analysis of adolescents Ģirts Burgmanis · Zaiga Krišjāne · Jurģis Šķilters
Received: 31 March 2013 / Accepted: 10 March 2014 © Marta Olivetti Belardinelli and Springer-Verlag Berlin Heidelberg 2014
Abstract We herein explore the perception of the geographic environment and analyse the mechanisms that constrain the cognitive processing of spatial information in general. Our guiding theoretical background assumption is that the structure of the spatial environment is a cognitively robust and mutually constrained threefold system relating (1) cognitive topology (comprised of a path and place structure of spatial information and constrained by reference frame-based factors), (2) experience-based functional knowledge (including the effects of socio-economic factors, frequency and familiarity) and (3) linguistic representations (primarily encoded in the prepositional system of a natural language). Here, we focus on (2), i.e. the effect of functional knowledge on the process of acquiring spatial knowledge. We empirically tested adolescents aged 12–17 years to explore the interaction between frequency, familiarity and functional knowledge from a developmental point of view. The social factors we explore are precisely defined and parameterized in our results (exposure to a particular urban area, place of residence, gender, age and factors relating to the environmental and social quality of the local area). Our research shows that there are divergences between the socalled objective topology and the cognitive typology of the urban environment that are significantly constrained by intensity of interactions with environment, number of functionally significant places within particular area and
Ģ. Burgmanis (&) · Z. Krišjāne Faculty of Geography and Earth Sciences, University of Latvia, Alberta Street 10, Riga 1010, Latvia e-mail: [email protected] J. Sˇk¸ilters Center for Cognitive Sciences and Semantics, University of Latvia, Riga, Latvia
age from a developmental perspective in terms of spatial knowledge acquisition. Keywords Spatial knowledge · Spatial cognition · Urban environment · Anchor points · Social value mapping
Introduction The processing of spatial information or, using the term of Downs and Stea (1973), ‘cognitive mapping’ through perception develops dynamic mental representations and transformations of the real and physical world, enabling humans mentally to structure a variety of place types and space in general. Although cognitive mapping allows an individual to build a coherent cognitive structure of an environment, the formation of this structure is constrained by three factors that interact with each other and are inherent in the spatial cognition: the frame of reference in which that individual is located (Coventry and Garrod 2004; Herskovits 1986; Landau and Jackendoff 1993; Zwarts 2005), the functional knowledge and individual acquires and uses in every day spatial processes (Coventry and Garrod 2004) and the language that is used in communicating their spatial experience (Coventry and Garrod 2004; Herskovits 1986). In the present study, we focus on the second system mentioned above, that of functional knowledge assuming that such spatial constraint is among the factors that determine the transformation of the physical space into the cognitive. We elaborate the variety of functional knowledge effects that determine spatial cognition in adolescents. In particular, we focus on the effects of frequency and familiarity as well as urban form on the acquisition of spatial knowledge within urban environment.
123
Cogn Process
Acquisition of spatial knowledge Studies on human spatial cognition mostly support the idea that the extent of spatial knowledge increases with the increase in familiarity with a particular area (Golledge 1978; Golledge and Spector 1978; Montello 1998; Siegel and White 1975). In other words, humans accumulate spatial knowledge through their spatial experience of acting in and interacting with the environment. This in turn means that familiarity determines not only the quantity, but also the quality of spatial knowledge about locations, distances and directions on various scales. Recent studies show that the spatial cognition is shaped by interactions of spatial and social knowledge (Maddox et al. 2008; Taylor et al. 2011). Therefore, the quality of spatial knowledge determined also by acquired information on social attributes of space including values, meanings, social interactions and peoples. The classical theoretical fundamentals of acquisition of spatial knowledge in large-scale environments are developed by Siegel and White (1975). These authors propose that internal representations of spatial knowledge of a place develop in three hierarchical stages from landmark recognition and development of knowledge about paths between landmarks to survey knowledge. In contrast to the theoretical framework put forward by Siegel and White, Montello’s (1998) ‘continuous framework of spatial microgenesis’ focuses on cognitive mapping as a continuous process and draws attention to the importance of familiarity and to individual differences in the acquisition of spatial knowledge and the generation of spatial representations. Spatial knowledge accumulation or increase in familiarity has progressive nature which is instantiated in the quantity, accuracy and completeness of spatial knowledge as an individual’s link with a place becomes stronger. Montello (1998) also proposed that individual differences may exist in the acquisition of spatial knowledge and that socio-demographic characteristics of individuals can determine the process of cognitive mapping.
Cognition of the urban environment The first noteworthy contribution to research on urban cognition was made by architects and urban planners who attempted to improve the visual attraction and functionality of the built environment (Appleyard 1973; Lynch 1960). Hence, the exploration of learning experience of urban environment may contribute and improve the basis of knowledge in city planning grounded by prominent urban planners such as Kevin Lynch. The acquisition of spatial knowledge in urban areas depends on an individual’s motion through the
123
environment and is based not only on direct perception, but more often relies on a learning experience (Lynch 1960). Therefore, it is clear that familiarity, which plays a crucial role in the cognition of an urban environment, determines that those objects and elements of an environment that occur more frequently will be more accessible in the human reference system and thus more easily identified later on. Golledge (1978) and Golledge and Spector (1978) drew on this idea of an important connection between motion, acts of navigation and familiarity with a particular urban area to propose the ‘anchor point’ theory. In this theoretical framework, the principal focus is assigned to the relative significance of landmarks, locations, paths and areas in the spatial environment for each individual. Initially, every individual has anchor points or primary nodes such as home, work, school and shops; all of which are important in his or her daily routine. Subsequently, an individual connects those anchor points or places with paths. An essential part of developing spatial knowledge is the search for paths and the journey between anchor points through segments of space. In the process of searching, an individual acquires knowledge and learns about new, oftenvisited places and landmarks defined as secondary, tertiary and lower-order nodes (Golledge and Stimson 1997). As the result of interaction along the paths between the nodes, there occurs a development of areal concepts such as neighbourhood, community and region. Couclelis et al. (1987) extend anchor point theory by suggesting that nodes are not always point-like (landmarks), but may be also be larger-scale areas and paths. Following the ideas proposed by Lynch and by Golledge, numerous studies have confirmed that landmarks as well as other categories of the urban environment (paths, edges, districts and nodes as proposed by Lynch 1960) are significant in learning, structuring and way-finding in urban areas (Aragones and Arredondo 1985; Golledge et al. 1993; Loomis et al. 1999; Wang and Spelke 2000). However, despite the influence and strong tradition of applying Lynch’s categories in urban cognition studies, some recent work has taken a more critical view on this object-centred approach. First, Golledge (1999) reconsiders the urban elements that shape spatial representations of a city and suggests that they should be viewed in a more general way in geometric terms of points, lines, areas and surfaces. Further, several authors emphasize that the urban elements proposed by Lynch are largely oversimplified and instead have to be considered as dynamic and complex entities acquired through interaction with the environment and also includes contextual knowledge consisting of a large amount of information on spatial features, the spatial interaction between users, the spatial conditions of the surrounding area (Stevens 2006; Portugali 2011) and social
Cogn Process
knowledge (Golledge and Stimson 1997; Maddox et al. 2008; Taylor et al. 2011). Therefore, not only single attributes, but also the functionality, symbolic meaning and form of an urban area in which a particular object is located are essential for the development of urban cognition. This approach has been employed and partially confirmed by empirical studies that use virtually generated environments to facilitate the exploration of spatial behaviour in large-scale areas (Newman et al. 2007; Waller et al. 2003). These studies show that the acquisition of spatial knowledge in cities may depend on the urban form. These studies also indicate that knowledge about an urban area can be acquired as a spatial layout as a whole and that small changes to this layout may affect cognitive mapping. In this case, general visual information about and configuration of the urban area play a more significant role than the single properties of the area in the development of awareness about that area. Similarly, Montello and Sas (2006) suggest that urban form affects the acquisition of spatial knowledge, and they conclude that a more articulated space that is broken up into more different parts, such as a built environment with high structural diversity in terms of colour, design and complex layout, is more difficult to perceive and use for routine spatial tasks. The effect of the spatial configuration of area on spatial urban cognition partially explains the ‘space syntax’ theory of Hillier (2012), which posits that there exist general spatial laws in terms of a human’s geometric intuition that are encoded or embedded through planning in the design of city and that at the same time allow an individual to conceive attributes of the urban environment. The line is the most appropriate element for human geometric intuition in that it preserves informational stability for moving individuals as well as visibility and an almost complete picture of an overall system (Hillier 2002). For these reasons, the grid system of street networks (i.e. containing areas with a clear geometry) allows patterns of urban form to be more easily and intuitively perceived and used for movement, as opposed to irregular, asymmetric and more complex urban layouts. This assumption also supported by several empirical works (Wang and Spelke 2002; Yaski et al. 2011).
Factors influencing children’s spatial cognition of the geographic environment Extensive research on children’s cognition and spatial behaviour started in the 1960s with a significant body of work by Jean Piaget (see for example Piaget and Inhelder 1967) that influenced the emergence of cognitive developmental studies in the field of human geography exploring the spatial abilities of young children aged 3–6 years old
(Blaut and Stea 1971; Blaut et al. 2003) and assessing their development of spatial knowledge in a large-scale environment (Hazen 1982; Hazen et al. 1978). Other studies have highlighted the acquisition of spatial knowledge by children aged 5–13 years from many different points of view (Golledge et al. 1985; Matthews 1987, 1992; Rissotto and Tonucci 2002, 2006). Matthews (1992) summarizes previous empirical evidence and indicates that three main factors affect children’s environmental cognition: active and passive exploration of the environment, the nature of activity and the level of activity control. More active and less controlled exploration of the accessible environment leads to more spatial representations. Matthews (1987) indicates that age and gender have an effect on the level of children’s spatial experience. By assessing free recall sketching, map interpretation and aerial photograph interpretation he found that the size of the ‘free home range’ of children aged 6–11 years is directly linked with their knowledge about the environment. Children aged 6–9 years and girls, who have stronger spatial restrictions, showed weaker awareness of the environment near to their place of residence. Similar results were found by Golledge et al. (1985), who used the route navigation of children aged 9–11 between two points as the basis for their research. The authors observed that children who are residents of the same neighbourhood have different levels of spatial knowledge that depend on the number of major nodes in the children’s route. Younger children with a limited free range in their residential neighbourhood cannot identify as many locations as those who have a larger daily living space. Also, children recognize a greater number of familiar locations more quickly and more accurately if they are experienced daily. Several empirical studies have extended the research on gender differences in the mental representations of space by focusing on the effect of children’s mode of travel to activity sites and the level of independent mobility on the development of spatial cognition. Garcia-Mira and Goluboff (2005) reported that there are different ways of conceptualizing the same space that depend not only on the age and gender of the children, but also on the mode of transport used. Children who walk to school acquire more information about their environment, remember more places and objects and demonstrate a greater ability to structure space. Rissotto and Tonucci (2002) explored independent/dependent mobility and confirm the importance of an active individual—the level of interaction with the environment acts as a constraining factor in children’s acquisition of environmental knowledge. They found differences in the extent of spatial knowledge that can be associated with the mode of children’s travel behaviour. Children who travel to school independently had better
123
Cogn Process
scores in the tasks used in the study (drawing a sketch map, reading a blank map and drawing a home to school itinerary map). Despite the long-standing tradition in cognitive geography and developmental psychology of focusing on exploring how young children acquire spatial knowledge, in our study we examine children aged 12–17 years. There are four key reasons for our choice of this particular age group. First, although children’s understanding and representations of space as well as their behaviour in space are of great interest to geographers, psychologists, cognitive scientists and environmental planners, there are not many studies that consider children’s acquisition of spatial knowledge in large-scale environments between the ages of 12 and 17. Second, from the development perspective, this age group corresponds to a transitional stage of human development— adolescence—when substantial socio-psychological changes (e.g. in the development of identity) occur, as do major changes in spatial behaviour through an increase in spatial autonomy. For instance, the spatial restrictions established by parents for growing adolescents gradually decrease, and the mode of mobility becomes more independent (O’Brien et al. 2000). Hence, this is the period in human life when a substantial transition from a conceptual to more functional knowledge of large-scale space occurs. Third, following Sibley (1995), we agree with the proposition that adolescents are the most spatially marginalized group in the modern urban environment. Hence, we would argue that it is crucial that we explore how the various modes and intensity of environmental interaction, as well as the daily experienced and accessible functionally and physically diverse urban environments, affect an adolescent’s spatial cognition of the geographic environment, in order to show the influence of modern urban life on adolescents’ cognitive development in general terms. Finally, we consider that the age group of 12–17-yearold adolescents may be a very useful and indeed more valuable source of data than older age groups for explaining divergences between the so-called objective topology of the environment and the cognitive topology of the environment, as well as eliciting the factors that may constrain the accurate perception of a real-world environment. This assumption is based on the thesis that children and adolescents mostly have not experienced significant transformations of the environment (e.g. change of place of residence, less travel experience).
Social value mapping technique Instead of traditional tools used for exploring external representations of stored spatial knowledge such as sketch
123
maps, map interpretation and verbal descriptions, we used the social value mapping technique (Tyrvainen et al. 2007) as our main tool for measuring the extent of respondents’ spatial knowledge. We consider the social value mapping task to be an appropriate method for exploring adolescents’ knowledge about space and their perception of the environment as well as the way in which the perceived real world is transformed into a cognitive format and the factors that influence this transformation. This opinion is based on two assumptions. First, as it was mention above, the social knowledge of space including attributive values and attitudes interacts with spatial (geometrical and topological) knowledge and, thus, together generate spatial cognition (Golledge and Stimson 1997; Maddox et al. 2008; Taylor et al. 2011). The second is grounded in the conceptual understanding of ‘place’ in human geography where it is defined as a part of space with human-assigned meaning (see, e.g. Agnew 1987; Cresswell 2004; Tuan 1975, 1977). Consequently, humans assign meanings primarily to functionally relevant parts of experienced space (for a slightly different point of view cf. Coventry et al. 2010; Coventry and Garrod 2004) in large-scale environments. Furthermore, functionality of place in an objective geographic environment can be mapped cognitively faster than other information, Based on this notion and developmental characteristics of age groups, we assume that social values such as safe/unsafe, pleasant/unpleasant and freedom from adult supervision clearly representing functionality of place can all be important for adolescents when they choose places in an urban environment to go about their daily social lives. It is therefore logical that those parts of the urban environment with such additional and relevant information can be identified by adolescents as a ‘places’ and can be cognitively mapped faster than other places in a particular area. These crucial aspects of human cognition determine the strength of social value mapping techniques that allow the storage of spatial experience, knowledge of the respondent and the building of internal representations of the external environment. When respondents were asked to assign numbers for social values to places on a real map they had to (a) use their stored spatial knowledge about the local environment, (b) search mentally for the places in their cognitive map where functionality corresponds to the conceptual meanings of social values and (c) build an internal cognitive representation of the environment. The present study focuses on the exploration of functional knowledge factors that may constrain the acquisition of spatial knowledge within a city, an environment with high and continuously growing complexity in terms of its physical and socio-economic structure. Having considered previous studies (see, e.g. Couclelis et al. 1987; Matthews 1987, 1992; Montello and Sas 2006;
Cogn Process
Yaski et al. 2011) and fundamental theoretical frameworks (Golledge 1978; Montello 1998; Siegel and White 1975), we propose following hypothesis. First, we assume that socio-demographic differences, duration of living in a specific place of residence, frequency of travel and number of activities in a particular neighbourhood are significant factors that can affect an individual’s ability to acquire a diverse amount of spatial knowledge concerning locations in the vicinity of residence. Second, we also believe that, in addition to familiarity with a particular area, which is gradually formed according to the amount of interaction with the accessible environment, humans acquire spatial knowledge about an urban area as a spatial layout as a whole and the differences in the urban form may have an effect on the generation of spatial representations. Therefore, this study tests a variety of variables: (1) time of residence in a particular neighbourhood/urban area, (2) urban form of the area, (3) frequency and type of environment usage, (4) mode of travel between activity places and (5) age and gender with familiarity of environment.
Case study areas The present research was undertaken in three study areas located in different parts of the largest city in Latvia, Riga. The areas were chosen so that each represented specific type of urban environment was physically and functionally distinct. Each of these areas contains two or more neighbourhoods that have similar physical settings. The first area (Area 1) is located within the central part of the city and can be characterized as an extension of the city centre (which is informally called ‘the silent centre’). The territory of Area 1 occupies 6.5 km2 and contains around 15,000 inhabitants in total (Riga City Council 2008). The area has a medium-density built environment where residential houses are mixed with diverse commercial, industrial and public service buildings. Most of the houses were built at the beginning of the twentieth century and are five-storey buildings. However, in the 1960s, a number of new residential multi-storey houses were built, thus bringing in an additional number of residents from a different socio-economic and socio-demographic background. The monocentric spatial and functional structure of Riga is grounded in its historical growth, which means that Area 1 contains more functionally significant objects and places than the two other areas covered by this study. Area 1 also has a more dense, irregular street network. The second area (Area 2) is located on the outer edge of the city on the right bank of the river Daugava and comprises three neighbourhoods—Purvciems, Plavnieki and Dreilini. The urban environment of the neighbourhoods included in Area 2 is similar to the residential areas of
other post-soviet cities where wide expanses of large-scale housing estates rarely overlap with small areas of green space and mixed-use areas. The construction of buildings mainly took place between the 1970s and the 1990s, a period corresponding to Soviet-era urban planning, architectural and construction outlines and typified by a high density of 5–16-storey reinforced concrete apartment houses that make the physical and functional structure of the area markedly oppressive and homogenous. Area 2 is therefore also characterized by a shortage of prominent objects and places of cultural and historical significance as well as a less-dense and more regular network of major streets. Area 2 is enclosed by major roads, which are also seen as barriers that prevent people from freely reaching and using the local public and green spaces, thereby restricting the spatial experiences of younger people due to safety concerns. Area 2 has about 85,000 inhabitants and covers an area of 13 km2 (Riga City Council 2008). The third area (Area 3) is located in the western part of Riga on the left bank of the river Daugava, and in this study represents a mixed built and mixed land-use urban environment. Area 3 consists of three neighbourhoods— Zolitude, Pleskodale and Sampeteris—with a total area of about 12 km2, all significantly different in terms of urban form. This means that these three parts of Area 3 have different urban density and make-up. Area 3 mostly consists of an industrial zone and a small residential zone where wooden one-storey houses from the beginning of the twentieth century coexist with houses that are no higher than three storeys, which were built between the 1950s and the 1990s. The more homogeneous part of Area 3 is situated in the south and is dominated by detached family houses. A remarkable feature of the southern part of Area 3 is that between the clusters of one-to-three-storey detached houses built during the twentieth century, there are also green areas of various sizes. A typical residential zone of Riga is located in the northern part of Area 3. It is a highdensity territory with 6–16-storey apartment houses built in the 1980s. In this part, the commercial and public service buildings are distributed along the large-scale housing estates and close to the major roads. Approximately 19,000 out of the 30,000 inhabitants in Area 3 live in these largescale housing estates (Riga City Council 2008) in the northern part of this study area. The choice of Areas 2 and 3 was based on the fact that, in contrast to Area 1, the locations of these two areas largely restrict opportunities for residents to reach essential objects with diverse functionality and the city centre with ease. Our selection of three such diverse areas was based on the assumption that the different locations of the respondents’ places of residence in the various urban environments and their access to essential objects and the city centre may affect their generation of representations of space.
123
Cogn Process Table 1 Socio-demographic characteristics of sample (N = 449) Characteristics
Area 1
Area 2
Area 3
Age (SD)
14.4 (1.6)
14.1 (1.6)
14.6 (1.4)
Female
54
48
42
Male
46
52
58
Primary
83
84
81
Secondary
17
16
19
96
97
73
4
3
27
N = 111
N = 184
N = 154
Sex (%)
Educational level (%)
Housing type (%) Apartment Private house Total sample
Methods Participants In February 2010, questionnaires were distributed to 942 students aged 12–17 years in three randomly sampled urban schools in the three areas described above. The socio-demographic characteristics of the sample are given in Table 1. Students who completed the questionnaire but stated that they lived outside the study areas were eliminated from the analysis, which meant that 468 (50 %) of the questionnaires were initially included; however, 449 (96 %) of these were accurately completed by the respondents and considered valid. Materials and procedure Participants completed a questionnaire-based survey consisting of three parts from which general and site-specific information about their spatial experience was gathered. Previous studies have considered the role of individual differences in the acquisition of spatial knowledge. We therefore included questions asking respondents for general socio-demographic information (age, sex, education level, housing type). The second part of the questionnaire was designed to elicit information about the respondents’ spatial behaviour in their place of residence (mode of travel to school, length of living in place of residence, amount of leisure time spent within the area daily). Most of the acquired information showed the respondent’s exposure to a particular urban area. This set of questions was designed to test the role of familiarity in the respondents’ cognition of geographic environment as discussed in several classical studies (Montello 1998; Siegel and White 1975). In the third part of questionnaire, we included two questions that would enable us to examine the anchor point hypothesis of Golledge (1978) and the importance of functionally
123
significant places in urban spatial cognition. We asked respondents to indicate on a map of the area the locations of (1) the place of residence of their best friend and (2) their main place of out-of-school activity. If these places were outside the area shown on the map or did not exist, the respondent ticked the box ‘not in this area’. These questions were included because we suppose that, besides home and school, these two places are the most important anchor points in the life of an adolescent and thus including them in our analysis would allow us to examine the effect of functionally significant places in the acquisition of spatial knowledge of this specific age group. The fourth part of the survey included questions aimed at capturing adolescents’ spatial knowledge of the area surrounding their place of residence by eliciting specific information about the particular social values attached to personally significant places. For our social value mapping task, which was originally a GIS-based method developed for spatial planning, we used 1:13,500 scale maps of the areas around the schools involved in the survey. The social values of places were created as conceptual constructs for the survey. This meant that the researchers had to formulate concepts of the environmental qualities of places that were then included in the questionnaires given to the respondents (Makinen and Tyrvainen 2008; Tyrvainen et al. 2007). Unlike the questionnaire in Tyrvainen et al. (2007), where the respondents were asked to assign an ordinal number of an area from the map of the case study area to a particular social value (e.g. a lovely view), we took a slightly different approach in that our respondents were asked to assign independently an ordinal number of social value to those places that in their opinion corresponded to the given value descriptions. Specifically, the respondents were asked to indicate on the map of their area of residence those places they considered pleasant, unpleasant, safe, unsafe and where they felt free from adult supervision (assigned ordinal numbers of social value to particular places as 1–5). There was also a limitation—only one place could be given one of the above five values. If the place the respondent had in mind was not on the map or they were unsure that such a place existed, they ticked the box ‘not in this area’ or ‘hard to say’. Data analysis In our analysis, we coded eight independent variables (sex, age, length of living in place of residence, travel mode to school, best friend’s place of residence within area, main place of out-of-school activity within an area, child’s level of independent mobility and amount of leisure time spent within the area daily), all of which we assume hypothetically may affect a human’s perception of their local urban environment. Statistical analysis of data was performed for
Cogn Process Table 2 Respondents’ spatial knowledge of the urban area surrounding their place of residence and school Independent variables
Area 1
Area 2 H
df
M
Area 3
M
SD
Girl
.54
.37
Boy
.54
.35
12–13
.44
.36
.47
.35
.42
.39
14–15
.54
.35
.41
.39
.50
.36
16–17 .68 .34 Duration of living in place of residence
.41
.38
.58
.37
.42
.38
.53
.39
.43
.37
.51
.36
SD
H
df
M
SD
H
df
.47
.39
1.188
1
.54
.35 5.49
2
.07
1
1.45
1
8.72**
1
.66
1
.516
2
Sex .01
1
.41
.40
.44
.34
.75
1
Age (years)
Less than 5 years
.39
.32
More than 5 years
.63
.35
Active
.56
.36
Passive
.52
.36
7.23*
11.08**
2
1
.41
.02
2
1
Travel mode to school .31
1
.43
.40
.43
.31
.10
1
.49
.38
.57
.34
Best friend’s place of residence within area Yes
.61
.34
No
.47
.36
4.08*
1
.49
.36
.29
.37
12.04**
1
.57
.36
.37
.35
Place of out-of-school formal activity within area Yes
.55
.36
No
.54
.36
.06
1
.32
.31
.45
.38
2.20
1
.55
.38
.50
.36
Child’s level of independent mobility (travelling alone to school) Low
.58
.39
Average High
.45 .55
.29 .36
1.13
2
.46
.36
.49 .42
.31 .38
1.38
2
.53
.33
.56 .50
.38 .38
Time spent on leisure activities within area Less than 2 h
.46
.37
.35
.35
.42
.39
2–8 h
.50
.37
6.82*
2
.45
.38
1.84
2
.50
.38
More than 8 h
.69
.30
.47
.37
.58
.37
2.98
2
Mean scores of the spatial knowledge index. Scores range from 0 to 1 Statistically significant at ** p \ .01; * p \ .05
each study area separately. The main aim of such an approach was to assess the potential effect of the area’s environment including location and physical and social setting as well as other factors that may have an influence on a human’s acquisition of spatial information. To study the factors influencing environmental cognition, we designed an additive index of spatial knowledge that included five variables obtained from the social value mapping task, namely pleasant (1), unpleasant (2), safe (3), unsafe (4) and areas where respondents feel free from adult supervision (5). The questionnaire answers were given 1 point if a value number (1–5) was assigned to place marked on the map. If a social value number was not linked to a place on the map and the box ‘not in this area’ or ‘hard to say’ was ticked, the answer was given a score of 0. Each respondent could thus accumulate a maximum score of five points for this aspect of the questionnaire. Next, for each
individual, we calculated the additive index of spatial knowledge (mean value total points). Thus, the level of each individual’s spatial knowledge could range between 0 and 1. The acquired index was used as the dependent variable in further analysis. The purpose of constructing a spatial knowledge index was to determine whether or not individual differences were shaped by adolescents’ sociodemographic background, their distinct spatial behaviour, or their functional interaction with the urban area have an affect on their amount of spatial knowledge.
Results and discussion The influence of background characteristics as independent variable-based differences in the mean values of the additive index of spatial knowledge was confirmed by the
123
Cogn Process
Kruskal–Wallis test (see Table 2). Table 2 illustrates the differences in the extent of the adolescents’ knowledge about the area surrounding their place of residence and school only. Foremost, we tested the hypothesis stating that (a) sociodemographic differences, (b) duration of living in a specific place of residence, (c) frequency of travel and number of activities in a particular neighbourhood, (d) factors representing familiarity have effect on individual’s ability to acquire a spatial knowledge concerning locations in the vicinity of residence. From Table 2, it can be seen that not all the independent variables that are statistically significant in terms of differences in the mean values of the additive index of spatial knowledge are appropriate for all studied areas. In Area 1, four factors (age, duration of living in place of residence, best friend’s place of residence within area and amount of leisure time spent on leisure activities within area) show stage progressing changes of spatial knowledge. In other two areas characterized as urban environments with a homogeneous physical spatial setting, only the presence of best friend’s place of residence within area influences adolescent’s extent of spatial knowledge. Although age as a factor affecting knowledge about space has usually been considered in studies on young children (Matthews 1987), we show that age also plays a role in the case of adolescents. We find significantly progressive differences in the number of places indicated by respondents, which is reflected in fluctuations in the additive index of spatial knowledge. We clearly distinguish differences between three age groups (12–13 year olds, 14– 15 year olds and 16–17 year olds) and therefore conclude that, like younger children, adolescents also acquire more knowledge about space with age. However, we would argue that the changes in the additive index of spatial knowledge can be attributed not only to increasing knowledge about space as such, but also to the developmental peculiarities of each individual and the gradual change in the way in which adults perceive the surrounding environment. An important factor that affected the adolescents’ spatial experience and increasing familiarity with their local urban environment was the duration of living in the place of residence. Respondents who had lived longer in the neighbourhood could identify a significantly greater number of the places mentioned in the questionnaire. This indicates that the cognitive map of the neighbourhood for this group is developed in more detail than for those who had moved to the area more recently. The results show that it is possible to identify an approximate time boundary beyond which the spatial experience of the individual becomes sufficient for fixing a cognitive topology and representing it in the perception of certain places within the area of residence. The results also show that, in our study,
123
changes of perception due to the accumulation of spatial knowledge are observable after 5 years of living in a particular place of residence as supposedly the territory becomes more familiar. Another important factor that reflects familiarity with the local environment is the amount of leisure time an individual spends in the area of residence outside school and home. Our results show that the more time an individual spends outside home and school, the better their knowledge about the respective territory, thus affecting the quantity and quality of the experience that the individual has of the surrounding territory. Those respondents who spent much of their time exploring the surrounding territory were better able to identify places and rate them according to social values. The location of the place of residence of the respondent’s best friend in the same area as the respondent was included as an independent variable in our analysis based on two considerations. First, we assume that the friend’s place of residence acts as another functionally significant and meaningful place—as an anchor point or primary node —in the adolescent’s objective topology of the local environment and also in the cognitive topology of that environment. The second consideration emerges from the first, namely that the presence of an additional meaningful place or node in the local environment is expected to increase the respondent’s familiarity with that environment. We offer the following explanations for this assumption. The increase in familiarity is associated with the habit of adolescents of going home from school in the company of a friend or friends. This habit can provide the adolescent with an opportunity regularly to visit additional places in the local area, varying the route based on their friends’ important places. Also, the relatively short distance between the respondent’s place of residence and that of their best friend (mostly within 1 km) allows them to reach each other easily and may be regarded as another explanation for the growing adolescent’s greater familiarity with their area. More frequent meetings provide adolescents with more opportunities to spend leisure time together and explore and experience the local environment, which facilitates the acquisition of more detailed spatial knowledge. Evidence to support our assumption is provided by the significant differences in the mean values of the additive index of spatial knowledge between respondents whose friends live in the same area and respondents whose friends do not. It is clear that the presence of the best friend’s place of residence in the same area is a very influential factor in an adolescent’s development of a more accurate cognitive topology and also applies in the case of areas with a spatially homogeneous physical setting (Areas 2 and 3).
Cogn Process Table 3 Relationship between spatial knowledge index and background variables: Spearman rank-order correlation coefficient (r) Independent variables
Area 1 rs
Sex
−.01
Area 2 rs
Area 3 rs
.064
.086
Age
.26**
−.083
−.133
Duration of living in place of residence
.32**
.010
−.022
Travel mode to school
.05
−.023
−.096
Best friend’s place of residence within area
.19*
Place of out-of-school formal activity within area Child’s level of independent mobility (travelling alone to school)
.02
−.109
.065
−.01
−.083
−.047
.087
.139
Time spent on leisure activities within area
.24*
.256**
.236**
Statistically significant at ** p \ .01; * p \ .05
Next, Spearman rank-order correlation analysis was used to identify the relationship between the spatial knowledge index and the background/independent variables. The relatively significant correlations presented in Table 3 point to correlations between the above mentioned four factors of age, duration of living in place of residence, best friend’s place of residence within area and amount of leisure time spent within the area daily, and the extent of spatial knowledge the individual has about the areas of residence. Only one of these four factors, namely best friend’s place of residence within area, has a significant effect on the increase in the respondents’ spatial knowledge in Areas 2 and 3. The correlation values in Table 3 show that with the increase in time that individuals had lived in the area and in time they had spent in the vicinity, there was a corresponding increase in their spatial knowledge of the environment. The increase in knowledge can be explained by a progressive increase in familiarity with the space— identification of routes, physical objects and peculiarities of the environment. Conversely, adolescents with limited leisure time spent in the territory of residence or a shorter duration of living in the neighbourhood showed poorer knowledge in the social value mapping task. These two factors must therefore be viewed as important constraints in terms of the adolescent individual’s perception of the environment and spatial representations. The correlations in Table 3 also show that older adolescents (16–17 years old) have a better knowledge of the area and are more fixed in the social values they assign to these places than the younger ones. The findings also show that in this age group, the influence of gender on the amount of spatial awareness is not significant.
Finally, using the previously mentioned correlations, we tested hypothesis stating that (a) humans acquire spatial knowledge about an urban area as a spatial layout as a whole and (b) the differences in the urban form may have an effect on the generation of spatial representations. The presented correlations once again confirm the importance of the presence of an additional functional anchor point or node for the acquisition of spatial knowledge in an urban setting that is homogeneous in terms of the built environment and has few functionally significant objects. This type of urban environment is observable in the majority of Riga’s residential areas and, in the case of this study, is represented by Areas 2 and 3 in Table 3. The results of the correlation clearly show that there exists a relationship between the urban form of the research area and the formation by the respondents of a cognitive topology of the environment of the area in which they live. We consider that this relationship can be explained by the spatial functional structure of the city of Riga, which is monocentric in form. The distance from the place of residence to the urban core plays an important role in the acquisition of spatial knowledge; the closer the adolescent’s place of residence is to the city centre, the more functionally significant objects are found in the immediate area and located close to each other, and the more dense and irregular the major street networks are. Consequently, due to this spatial pattern of objects, the closer to the city centre the individual lives, the more time they need to acquire spatial knowledge of those objects and the more nodes or meaningful places those individuals may have in the vicinity of their place of residence.
General discussion Our study focused on the ways how functional knowledge (Coventry and Garrod 2004) (as linked to frequency, familiarity and urban form) constrains and determines the transformation of physical space into cognitive one. The use of social value mapping techniques (Tyrvainen et al. 2007) has provided evidence in our study on how an individual modifies the physical urban environment into a cognitive map and how this depends on factors such as familiarity with a particular urban environment and age. First, our findings support our assumption that the duration of living in the place of residence and the amount of leisure time spent in the area are directly associated with the amount of acquired spatial knowledge. An increase in an adolescent’s age also leads to an increase in their familiarity with the surrounding environment and overall knowledge of spatial relationships. These findings are particularly consistent with the theoretical approach of
123
Cogn Process
Montello (1998), who emphasizes that the acquisition of spatial knowledge about a particular area is a continuous process in human cognitive development. The continuous framework of spatial microgenesis posited by Montello (1998) and in other previous research (O’Brien et al. 2000) also explains continuous changes in the amount of spatial knowledge in the case of adolescents. The extent of spatial experience and therefore amount of spatial representations also continue to increase after the age of 12 due to gradually decreasing restrictions and an increasing amount of independent mobility. We suppose that this result is directly associated with the significant increase in individual differences and personality features, which develop with age and which are inevitably present in an individual’s perception of their environment. The presence of the place of residence of a best friend is indisputably a very strong factor that affects the amount of spatial knowledge, which in turn indirectly affects familiarity with a particular area. Anchor point theory (Golledge 1978) can provide an explanation for this correlation in our study in that Golledge found that the existence of an additional node in the local environment has a significant effect on the formation of cognitive topology (a finding which is also consistent with Golledge et al. 1985). Second, the results of our analysis of the impact of the spatial structure of different urban areas on the acquisition of spatial knowledge prove our hypothesis that the physical form and structure of the urban environment exert a constraining effect on the acquisition of spatial knowledge and the formation of spatial cognition. This finding is partly consistent with previous evidence that supports the opinion that the whole layout of an urban area may influence the acquisition and use of spatial knowledge (Montello and Sas 2006; Yaski et al. 2011). The present study was performed in various three different parts of a city that has a monocentric functional spatial structure, and our findings demonstrate that, in contrast to areas located within the urban core that have a more heterogeneous physical and functional structure as well as a more dense and irregular street network (Area 1), urban areas without notable landmarks and with a homogeneous physical structure including a sparse and regular major street network (Area 2 and Area 3) do not encourage the progressive development of cognitive maps, and moreover partly limit the individual’s ability to generate spatial representations of their environment. Considering these findings regarding the three case study areas and referring to theoretical works of Lynch (1960), Golledge (1999) and Hillier (2012) that support the notion that urban forms generally are perceived as geometric configurations that are generally shaped by lines or paths that are reflected in the urban environment as a street network, we would argue that an urban area with a dense
123
and irregular street network is harder to understand than an urban area with a sparse, linear and regular layout. Furthermore, we would also suggest that, although a higher density of street network adds more complexity to the layout of an urban area, which makes it more difficult to know fully, at the same time, this inherent complexity allows the individual to generate more spatial representations over time. Hence, the more central the place of residence, the more gradual, continuous and complete is the cognitive mapping. To conclude, our research findings strongly suggest that the mapping of objective topology onto the cognitive is enhanced by familiarity with the environment as well as by familiarity with the functionally significant objects in a particular area. The more actively the environment is processed, the richer the representations generated. This finding supports prior research that suggests that active and independent exploration of large-scale environments is important in the acquisition of spatial knowledge (Matthews 1992). In our study, the longer the adolescents have resided in the area and the more frequently they spend leisure time there, the more places they can associate with the prescribed social values. Despite the difference in the nature of our study area (in our case, large-scale spatial environments), we believe our results support the findings of Coventry and Garrod (2004), which show that the underlying structure of spatial cognition is supported by a wide range of functional knowledge and co-influenced by socio-economic, cultural and other factors such as spatial behaviour and transportation mode that then determines the individual’s perception of geographic environment and acquisition of spatial knowledge. Acknowledgments The research was promoted with support of the European Union Social Fund project ‘Support of Doctoral Studies at the University of Latvia’, Project No. 2009/0138/1DP/1.1.2.1.2/09/ IPIA/VIAA/. Jurģis Šķilters’ work is supported by the Fulbright Scholar Program.
References Agnew J (1987) Place and politics: the geographical mediation of state and society. Allen & Unwin, Boston Appleyard D (1973) Notes on urban perception and knowledge. In: Downs R, Stea D (eds) Image and environment: cognitive mapping and spatial behavior. Aldine, Chicago, pp 109–114 Aragones JI, Arredondo JM (1985) Structure of urban cognitive maps. J Environ Psychol 5:197–212 Blaut JM, Stea D (1971) Studies of geographic learning. Ann Assoc Am Geogr 61:387–393 Blaut JM, Stea D, Spencer C, Blades M (2003) Mapping as a cultural and cognitive universal. Ann Assoc Am Geogr 93:165–185 Couclelis H, Golledge RG, Gale N, Tobler W (1987) Exploring the anchor-point hypothesis of spatial cognition. J Environ Psychol 7:99–122
Cogn Process Coventry KR, Garrod SC (2004) Saying, seeing and acting: the psychological semantics of spatial prepositions. Psychology Press, New York Coventry KR, Lynott D, Cangelosi A, Monrouxe L, Joyce D, Richardson DC (2010) Spatial language, visual attention, and perceptual simulation. Brain Lang 112:202–213 Cresswell T (2004) Place: short introduction. Blackwell, Malden Downs R, Stea D (1973) Image and environment: cognitive mapping and spatial behavior. Aldine, Chicago Garcia-Mira R, Goluboff M (2005) The perception of urban space from two different viewpoints: pedestrians and automobile passengers. In: Garcia-Mira R, Uzzell DL, Real JE, Romay J (eds) Housing, space and quality of life. Antony Rowe Ltd., Chippenham Golledge RG (1978) Learning about urban environments. In: Carlstein T, Parkes D, Thrift N (eds) Timing space and spacing time: making sense of time, vol 1. Edward Arnold, London, pp 76–98 Golledge RG (1999) Human wayWnding and cognitive maps. In: Golledge RG (ed) Wayfinding behavior: cognitive mapping and other spatial processes. The Johns Hopkins University Press, Baltimor, pp 5–45 Golledge RG, Spector AN (1978) Comprehending the urban environment: theory and practice. Geogr Anal 10:403–426 Golledge RG, Stimson RJ (1997) The spatial behaviour: a geographical perspective. The Gilfourd Press, New York Golledge RG, Smith T, Pellegrino JW, Doherty S, Marshall SP (1985) A conceptual model and empirical analysis of children’s acquisition of spatial knowledge. J Environ Psychol 5:125–152 Golledge RG, Ruggles AJ, Pelegrino JW, Gales ND (1993) Integrating route knowledge in an unfamiliar neighborhood along and across route experiments. J Environ Psychol 13:293–307 Hazen NL (1982) Spatial exploration and spatial knowledge: individual and developmental differences in very young children. Child Dev 53:826–833 Hazen NL, Lockman JJ, Pick HL Jr (1978) The development of children’s representations of large-scale environments. Child Dev 49:623–636 Herskovits A (1986) Language and spatial cognition. Cambridge University Press, Cambridge Hillier B (2002) A theory of the city as object: or, how spatial laws mediate the social construction of urban space. Urban Des Int 7:153–179 Hillier B (2012) Studying cities to learn about minds: some possible implications of space syntax for spatial cognition. Environ Plann B 39:12–32 Landau B, Jackendoff R (1993) “What” and “where” in spatial language and spatial cognition. Behav Brain Sci 16:217–265 Loomis JM, Klatzky RL, Golledge RG, Philbeck JW (1999) Human navigation by path integration. In: Golledge RG (ed) Wayfinding behavior: cognitive mapping and other spatial processes. The Johns Hopkins University Press, Baltimor, pp 125–151 Lynch K (1960) The image of the city. MIT Press, Cambridge Maddox KB, Rapp DN, Brion S, Taylor HA (2008) Social influences on spatial memory. Mem Cogn 36:479–494 Makinen K, Tyrvainen L (2008) Teenage experiences of public green spaces in suburban Helsinki. Urban For Urban Green 7:277–289 Matthews MH (1987) Gender, home range and environmental cognition. Trans Inst Br Geogr 12:43–56 Matthews MH (1992) Making sense of place: Children’s understanding of large-scale environments. Barnes & Nobles Books, Savage
Montello DR (1998) A new framework for understanding the acquisition of spatial knowledge in large-scale environments. In: Egenhofer MJ, Golledge RG (eds) Spatial and temporal reasoning in geographic information systems. Oxford University Press, New York, pp 143–145 Montello DR, Sas C (2006) Human factors of watfinding in navigation. In: Karwowski W (ed) International encyclopedia of ergonomics and human factors, 2nd edn. CRC Press, Taylor & Francis, Ltd, London, pp 2003–2008 Newman EL, Caplan JB, Kirschen MP, Korolev IO, Sekuler R, Kahana MJ (2007) Learning your way around town: how virtual taxicab drivers learn to use both layout and landmark information. Cognition 104:231–253 O’Brien M, Jones D, Sloan D, Rustin M (2000) Children’s independent spatial mobility in the urban public realm. Childhood 7:257–277 Piaget J, Inhelder B (1967) The child’s conception of cpace. W. W. Norton, New York Portugali J (2011) Complexity, cognition and the city. Springer, Hedelberg Riga City Council (2008) Statistical materials of Riga City Council City Development Departament. Riga City Council City Development Departament, Riga Rissotto A, Giuliani MV (2006) Learning neighborhood environments: the loss of experience in modern world. In: Spencer C, Blades M (eds) Children and their environments: learning, using and designing spaces. Cambridge University Press, Cambridge, pp 75–90 Rissotto A, Tonucci F (2002) Freedom of movement and environmental knowledge in elementary school children. J Environ Psychol 22:65–77 Sibley D (1995) Geographies of exclusion: society and difference in West. Routledge, London Siegel AW, White SH (1975) The development of spatial representations of largescale environments. In: Reese HW (ed) Advances in child development and behavior, vol 10. Academic Press, New York, pp 9–55 Stevens Q (2006) The shape of urban experience: a reevaluation of Lynch’s five elements. Environ Plann B 33:803–823 Taylor HA, Wang Q, Gagnon SA, Maddox KB, Brunye´ TT (2011) The social connection in mental representations of space: explicit and implicit evidence. In: Egenhofer M et al (eds) COSIT 2011, LNCS 6899. Springer, Heidelberg, pp 231–244 Tuan Y-F (1975) Images and mental maps. Ann Assoc Am Geogr 65:205–213 Tuan Y-F (1977) Space and place: the perspective of experience. University of Minnesota Press, Minneapolis Tyrvainen L, Makinen K, Schipperijn J (2007) Tools for mapping social values of urban woodlands and other green areas. Landsc Urban Plan 79:5–19 Waller D, Loomis JM, Steck SD (2003) Inertial cues do not enhance knowledge of environmental layout. Psychon B Rev 10:987–993 Wang RF, Spelke ES (2000) Updating egocentric representations in human navigation. Cognition 77:215–250 Wang RF, Spelke ES (2002) Human spatial representation: insights from animals. Trends Cogn Sci 6:376–382 Yaski O, Portugali J, Eilam D (2011) City rats: insight from rat spatial behavior into human cognitionin urban environments. Anim Cogn 14:655–663 Zwarts J (2005) Prepositional aspect and the algebra of paths. Linguist Philos 28:739–779
123
