Content uploaded by Agnieszka Olszewska-Guizzo
Author content
All content in this area was uploaded by Agnieszka Olszewska-Guizzo on Nov 07, 2017
Content may be subject to copyright.
Article
What makes a landscape
contemplative?
Agnieszka A Olszewska and Paulo F Marques
University of Porto, Portugal
Robert L Ryan
University of Massachusetts, USA
Fernando Barbosa
University of Porto, Portugal
Abstract
For centuries, landscape architects, architects, and urban planners have been designing outdoor
green spaces for one to contemplate. In today’s urban realm, we can understand a contemplative
space more specifically as one joining esthetic and environmental values with mental health
benefits for its visitors. So far, the concept of contemplativeness of a space has not been
operationalized and a definitive list of design principles of such a space has not been
developed. In response to this gap of knowledge, we have identified a set of features that may
be used in order to design and create a space of contemplation within seven categories:
Landscape Layers, Landform, Vegetation, Light and Color, Compatibility, Archetypal Elements,
and a Character of Peace and Silence. The developed framework is based on development and
analysis of a Contemplative Landscape Questionnaire. This instrument was developed based on
literature review and Delphi expert evaluation of multiple landscapes. The statistical tests on the
Contemplative Landscape Questionnaire revealed satisfactory reliability and validity measures,
which provided evidence-based support for the efficacy of designed spaces. This approach could
enhance the practice of landscape architects and urban designers by reinforcing intuition-
based designs with scientific evidence. The developed framework can also serve to identify
contemplative spaces for subsequent research purposes.
Keywords
Contemplative landscapes, landscape architecture, parks, gardens, design methods, environmental
impact, perception, spatial analysis, urban design, environmental psychology
Introduction
Most invented green spaces are considered places for calm relaxation, reconnecting with
nature, and contemplation. Landscape architecture, as an art discipline, understands
Corresponding author:
Agnieszka A Olszewska, FCUP, University of Porto, Rua do Campo Alegre, 687, 4169-007 Porto, Portugal.
Email: olszewska.agn@gmail.com
Environment and Planning B:
Planning and Design
0(0) 1–19
!The Author(s) 2016
Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/0265813516660716
epb.sagepub.com
contemplation as a state of mind, where we direct our attention toward a piece of art, or in
this case, the designed landscape (Norberg-Schulz, 1976). Professionals who design or review
urban green spaces often use the word ‘‘contemplation’’ to describe a desired output or
function of these spaces. For example, the designer of the famous Porto City Park,
Sido
´nio Pardal, describes his creation: ‘‘The Park’s landscape is an end in itself and
expresses its essence. It doesn’t attempt to imitate nature and has no other purpose
than direct use as a public urban space for recreation activities and contemplation ...’’
(Pardal, 2006).
In fact, reflection on any invented landscape, in terms of the psychological or mental
states they induce, ultimately leads to the concept of contemplation. Contemplation is a
major part of passive recreation, including activities requiring a minimal use of facilities and
which have a low environmental impact on the recreational site, i.e. walking, hiking, bird
watching, and all sorts of sedentary activities. Therefore, contemplation can be associated
with exposing oneself to landscape scenes while simply relaxing and being in nature, which
seems to be the main motive for visiting the parks by urban dwellers (Chiesura, 2004).
Although technical information in the design professions for recreation and park facilities
is plentiful (Harris and Dines, 1988; Neufert and Neufert, 2012), there is very little on the
physical attributes that create a contemplative outdoor space.
In the existing literature about contemplative landscapes, authors mostly refer to
contemplation as an effect that some particular designed space had on them as visitors.
This type of literature does not follow the rigorousness of the scientific process and is
mostly based on subjective descriptions, reports, or summaries about the designed areas
from the expert’s point of view. This suggests that the described attributes of such spaces
are potentially contemplative. For example, Heinrich Hermann (2005) describes two case
studies of contemplative outdoor places: The Salk Institute in California and the Woodland
Cemetery in Stockholm. He provides a deep analysis of design attributes of each of those
spaces in association with reflections and other psychological effects that he experienced
during his visits, and assumingly can be experienced by anybody visiting those places.
Contemplation according to him is an inner silence ‘‘understood not only in the acoustic
sense, but as the absence of distracting sensual stimuli.’’ What is more, he suggests which
physical attributes of the designed space can be associated with the psychological response
labeled as contemplation (e.g., the visibility of shade movements along the daily cycle
symbolizing the transcendence of time by ‘‘making one acutely aware of the passage of
time and its cyclical nature’’ (Hermann, 2005: 46). Similar approaches have been made by
Krinke (2005) and Treib (2005) who described how particular case examples from the
existing world can create a feeling in the observer that can be referred to as
contemplation. Their essays were gathered in the book Contemporary Landscapes of
Contemplation, which proved to be a vital source of reference for this study (Krinke,
2005). Contributors to this book (including Marc Treib, Heinrich Hermann, Michael
Singer, among others) consistently define contemplation as a fixed attention, an
elimination of thought, an inner orientation of oneself, and the reaching of an inner
silence. In seeking after the set of physical attributes of contemplative spaces, they present
different case examples of spaces that, according to them, are the most contemplative (e.g.
Bloedal Reserve in Washington State, USA; National Library in Paris, France; Thermal
Baths in Vals, Switzerland; Lightning Field in New Mexico, USA, among others).
Another focus in the literature review is the concept of restorative environments that
potentially share similar values with the contemplative landscapes. The concept of a
restorative environment was introduced in the 1980s by environmental psychologists
Stephen and Rachel Kaplan (Herzog and Barnes, 1999; Kaplan and Kaplan, 1989,
2Environment and Planning B: Planning and Design 0(0)
Kaplan, 1995; Ulrich and Parsons, 1992). This concept distinguishes contact with natural
vegetation as the main contributing factor of the restorative experience, which improves
health and well-being by eliminating mental fatigue, provides recovery of directed
attention capacity, and enhances the ability to reflect on issues of importance (Herzog
et al., 2003). Even though environmental psychologists did not address contemplation in
their study, their theory can provide a useful tool for examining contemplative landscapes.
In their book The Experience of Nature: A Psychological Perspective, the Kaplan and Kaplan
(1989) describe four attributes of the restorative experience—fascination, being away, extent,
and compatibility. However, these are only partly formalized in the physical space (e.g.
wilderness experience as part of being away, densely vegetated landscapes with natural
character, etc.)
More recently, Stiggsdotter (2005) used the restoration theory to more precisely establish
the physical features of the space that improve visitor health and well-being. She describes
eight characteristics of a healthy (restorative) public garden, which include serene, wild, rich in
species, spacious, common, pleasurable, festive, and cultural. The description of those
characteristics provides some technical design guidelines (e.g. biodiversity, character of
peace and silence, or sense of wilderness achieved by low maintenance of some parts of the
garden). Similar to many of the articles in this area, this work appears to be based more on the
theorist’s or designer’s personal observations rather than empirically based research studies.
The literature also revealed a focus on design strategies for achieving particular effects in
space. For example, Skalski (2005), whose research interest focuses on the values of long
vistas in urbanized landscapes, pointed out several benefits of creating long-distance view in
cities, which can be associated with the contemplative experience of the space. These include
a ‘‘sense of personal freedom ...to use of natural, perceptional possibilities’’ and ‘‘mental
pleasure,’’ which all lead to stress reduction.
There is also extensive literature by theorists, geographers, and designers about the use of
symbols and archetypes in the designed space, which may have a strong influence on the
psychological response. Some theorists (e.g. Hermann, 2005; Tuan, 1974) refer to archetypal
elements as space elements that induce contemplative responses by triggering some
particular emotions and reflections. However, they do not list all of the possible
archetypes that one can encounter in parks. Instead, as previously mentioned, they
highlight an existing archetype in a described space and describe its importance. Due to
this, additional research on the symbolic and archetypal elements had to be
undertaken in order to find which ones can appear in the designed urban landscapes and
what is their significance. One important reference for establishing a list of landscape-
archetypal elements is Jung’s work on dream analysis and symbols (Jung, 1955, 1964).
Symbols are important elements of the collective unconscious for they have been present
in our civilization since prehistoric times (e.g. the big stone boulder has consistently been
associated with the grave and, due to this, has a potential to induce contemplative,
commemorative experiences).
The literature about contemplative spaces refers to different psychological benefits that
may be achieved by visiting and emerging in such spaces, but there does not exist one
established list of physical attributes that would make it contemplative. Also, not every
author uses the word ‘‘contemplation’’ to describe the possible psychological benefits on
visitors induced by a particular space. This makes it challenging to determine if a particular
study is relevant or applicable to the study of contemplative landscapes or if they are
addressing some other concept. Therefore, there is a need to study the concept of
contemplative landscapes, particularly with a focus on operationalizing this term within a
rigorous, empirical study.
Olszewska et al. 3
Due to the fact that concept of contemplative space has a broad range of potential
meanings this study associates it with all psychological benefits that green designed spaces
can offer in connection to passive park recreation (e.g. a sense of well-being, inner silence,
relaxation, focused attention, attention restoration, stress reduction, recovering of
mental fatigue). By looking at how professional landscape architects, namely experts,
describe the physical attributes of a contemplative space, we sought out a common
understanding of such a construct and developed a framework to operationalize its
evaluation in designed settings.
Recently, there has been a strong trend in landscape architecture for evidence-based
design. Clients, including the general public and government agencies, are increasingly
putting pressure on designers to demonstrate that their projects will achieve both the
environmental and social benefits proposed. Mental health and well-being are more and
more often among the main targets of such attempts (Chiesura, 2004; Eberhard, 2009;
WHO, 2005; Windhager, 2009). Therefore, there is a clear interest in new solutions that
could help reaching these goals.
In summary, it would be useful to operationalize the artistic concept of contemplation.
This would allow researchers to analyze contemplative landscapes using scientific
approaches, by comparing them with noncontemplative ones in terms of their impacts on
individuals for example. There is at least one controlled study examining brainwave patterns
induced by contemplative landscapes (Olszewska, 2015). The results of this study suggested
greater activation in the brain regions responsible for visual attention when participants were
viewing landscapes which scored highest on the Contemplative Landscape Questionnaire
(CLQ). This suggests that contemplative landscapes can stimulate the benefits associated
with switching the mechanisms of attention. Nevertheless, this is only one example of how
the operationalized concept of contemplative landscapes can be used by researchers, and
why it is worthy to design and experience contemplative landscapes in our cities.
Methods
In order to come up with the range of elements and strategies that designers may use to
create contemplative spaces, we used methods well known in landscape architecture
research, including:
1. Literature review on contemplative landscape features
2. Visual Resources Management (VRM) model of evaluating the scenic quality of natural
landscape settings (U.S.D.A., Forest Service, 1973)
3. Delphi method, based on the evaluation by a panel of experts who form a consensus
about the problem with multiple possible scenarios (Hsu and Sanford, 2007)
4. On-site Photographic Documentation and Analysis (Markwell, 2000)
5. Sampling method according to principles of environmental photograph-based research
(Kaplan et al., 1972; Kaplan and Kaplan, 1989)
1. Literature review:
In order to develop a questionnaire to study contemplative landscapes, we used a review
of previous research to develop a checklist of all possible physical characteristics of the
contemplative spaces. All design strategies, tips, and attributes present in already existing
contemplative places were listed and described with the literature reference (Table 1), the
output was a 37-item checklist.
4Environment and Planning B: Planning and Design 0(0)
Table 1. Checklist of contemplative landscape characteristics—37 items characteristic for the
contemplative landscapes with the source references from the literature.
Physical attributes
SourcesItem # Design strategies
1 Long-distance view (400 m) 1, 6, 8, 10
2 Historic and ideological universalism 7
3 Inward orientation of the spatial composition 2, 4
4 Large space of absence (clearing) 3, 2
5 Smooth land form (mounds) 2, 8
6 Signage elements in the hierarchical relation to each other (central/
peripheral)
2, 8
7 Natural asymmetry (organic forms) 2, 10
8 Contrast with surrounding urban landscape (enclosure or visual distinction) 3, 2
9 Openings and closings of views 2, 8, 10
10 Spatial order, harmony, absence of disturbing stimuli 3, 2, 8, 10
11 Physical and visual relations are worked out 2, 8
12 Lack of direct exposition to the sun (point of view in shaded area) 2, 4
13 Warm, broken colors 2
14 stimulation to look up to the sky (e.g. optical manipulation of the skyline,
water mirror)
2, 11, 12
15 Simplification of forms 3, 2, 4
16 Repetition 4, 12
17 Visibility of shade movements along the daily cycle; sun/moon passage 2, 4, 10
18 Seasonally changing vegetation 2, 10
19 Character of peace and silence 2, 5, 13
20 High degree of wilderness 5, 9, 10
21 Biodiversity (rich in species) 5, 9
22 Large-scaled elements in relation to human body 2, 14, 10
23 Signage or symbolic elements with strong connotation to life and death 2, 10
24 Signage or symbolic elements with strong connotation to continuity of life,
collective human family
2, 10
Archetypal elements 2, 10
25 Path 3
26 Still water 15, 16
27 Waterfall 15
28 Single old tree 16
29 Big stone 15
30 Clearing 3
31 Forest 3, 16
32 Grave 16
33 Circle 16
Psychological elements
34 Gives sense of solitude 2, 5, 13
35 Invites to rest and relax 2, 5
36 Gives reorientation (from everyday routine, banality of life) 2, 5, 13, 10
Landscape composition type
37 Canopied, enclosed, focal, feature, panoramic 1
Notes: (1) Smardon et al. (1986), (2) Hermann (2005), (3) Krinke (2005), (4) Treib (2005), (5) Stigsdotter (2005), (6) Skalski
(2005), (7) Krippendorff (1995), (8) Kaplan et al. (1998), (9) Kaplan and Kaplan (1989), (10) Tuan (1974), (11) Hou (2015),
(12) Zelanski and Fisher (1996), (13) Herzog and Barnes (1999), (14) Bell (1998), (15) Jung (1955), and (16) Jung (1964).
Olszewska et al. 5
Twenty-four identified items were physical attributes of the space, nine constituted
archetypal elements originating from nature, three items focused on the psychological
effects on the visitor, and a final item served to identify the spatial composition type.
2. Building a contemplative landscape model (CLM) based on VRM:
Providing a clear, ‘‘rater-friendly’’ evaluation tool was the next step. Thus, the Checklist
was transformed into an online-based Questionnaire using the structure of the VRM model.
The VRM model, developed by the US Forest Services, is based on visual studies conducted
for the construction of utilities and other infrastructure on public lands throughout the
western United States (U.S.D.A. Forest Service, 1973). According to the VRM, all
landscapes have some level of scenic value that can be assessed. Although it is designed
more for natural landscapes (with some level of ‘‘cultural modifications’’) and is not too
sensitive to rural or urban landscapes, it has served as the basis for numerous landscape-
visual quality studies.
The decision to use the VRM model was motivated by a similar base assumption—each
given outdoor space has some level of contemplativeness that can be evaluated. The chart
below (Figure 1) presents this transformation process of the VRM model, which led to
creating a new CLM. The scheme also incorporates an example of the Boston model,
which is an adaptation of the VRM model used to evaluate the visual impact of wind
energy turbines and other new development in the Boston Harbor Islands National
Recreational Area (Burney et al., 2007).
Some categories of our contemplative model remained the same as in the VRM and
Boston models: Landform,Vegetation, and Adjacent Scenery (equivalent to Compatibility).
One very important key element of the contemplative landscape that the VRM model did
not consider is Layers of the Landscape (i.e. the depth of the view), which is connected
directly to the visibility of three planes and the comfort of long-distance views.
The VRM and Boston models incorporated Water as a key element in the visual quality
of landscape evaluation. However, each of these models has slightly different reasons for
addressing this element. In the VRM model, it is present because water is an important
factor of any natural landscape and generally increases the attractiveness of this landscape.
On the other hand, the Boston model is prepared for evaluating harbor landscapes, where
water is present in every evaluated case. In the CLM, water may be considered an archetypal
landscape element. Therefore, the Water element was fully transferred to the new category
Archetypal Elements, which replaces the Scarcity/Rarity categories. The scarcest elements of
landscapes according to the VRM model are, in fact, archetypal elements—with features
such as stones, waterfalls, and dense forests as strong archetypes. Thus, we decided to create
the all new category of Archetypal Elements, which combines both the water and scarcity
elements from the VRM model. Archetypal elements to consider include path, clearing,
single old tree, forest, still water (water mirror), waterfall, circle, grave, and boulder.
The VRM category Colors appeared incomplete for use in the contemplative model,
because most of the colors that we perceive depend on the light conditions of the site.
Also, many contemplative attributes of the space dealt with light and shade. Therefore,
the Color category was extended to the Color and Light category.
Finally, we created two all new categories that did not exist in either the VRM or the
Boston model, because these models dealt with more or less natural scenery. These new
categories are Artistic Expression and Character of Peace and Silence. Implementing these
elements was necessary in order to make the CLM sensitive to the designed landscapes in
urban areas. These were designed with artistic concern, mostly on a completely modified
6Environment and Planning B: Planning and Design 0(0)
environment, using some artistic expression strategies, including large-scaled elements in
relation to the human body, repetition, natural asymmetry, simplification of forms
(abstraction), architectural/geometrical use of plants. Also, the character of peace and
silence is very important for urban contemplative spaces, as it is a contrast to the noise
and crowdedness of the typical urban environment (in natural landscapes, this factor is
usually assumed to be the case).
Figure 1. Scheme of transformation of the VRM model, using inputs from the list of contemplative
landscape characteristics, to create the CLM—answer key for the questionnaire.
Olszewska et al. 7
After incorporating Checklist items within the VRM structure, a scoring system was
applied in order to create the scoring key to be used directly by the experts.
Scoring system. The CLM, as well as the VRM and Boston model scoring systems, was
based on a Likert scale, but with one major difference. Both the VRM and Boston models
used a 5-point Likert scale, while we opted for a 6-point Likert scale as to avoid the so-called
central tendency bias—a data collection error caused by the odd number of points in the
Likert scale, which creates a single central point and a propensity in the respondents to
choose it (Tatler, 2007).
3. Developing the questionnaire with the Delphi technique:
Asking experts was considered the best approach to check if a given item would
characterize a landscape as contemplative or not and then to develop a model based on
those findings. Asking the public about contemplativeness of space would be feasible, but
would also have more limitations than asking experts. Multiple items from the list presented
in Table 1 concern technical aspects of space design, because they originate from textbooks
written by experts and not directly addressed to the general public. The need to interpret
these technical terms for the public would dilute the nuances within the expert approach and
would require both additional knowledge and a detailed explanation, which would neither
be feasible within this study nor desirable from a methodological standpoint. Moreover, due
to the lack of technical criteria, variances on the appraisal of the landscape characteristics
would likely be higher, leading to the need of a larger sample size than in the case of expert
evaluation. Thus, the survey was addressed only to experts with the understanding that
subsequent phases of the research could include the general public once the CLM was
calibrated with expert responses.
The development of the questionnaire was based on the Delphi method, which is a
commonly used tool to establish a professional, objective judgment about a complex
problem with a wide range of scenarios. The evaluation of landscapes according to set
criteria fits well into that concept (Hsu and Sandford, 2007). According to the Delphi
method, experts (raters) should be carefully selected individuals with a background related
to the topic of investigation.
There is no consensus on the optimal number of subjects in the Delphi expert evaluation
(Hsu and Sandford, 2007). If the experts represent different areas, then up to 50 respondents
are recommended. If the background of the Delphi subjects is consistent, which is the case
for this study, 10–15 would be an optimal number of experts (Delbecq et al., 1975).
Therefore, 21 experts from around the world were invited to participate in our
investigation. The data collection was closed upon reaching 10 respondents, which was
two months after the initial invitation. All participating experts (four women and six
men) were academics in the area of landscape architecture at the PhD level with 6–31
years of professional practice and at least one peer-reviewed publication on landscape
design strategies, perception of the landscape, theory of design, or environmental
psychology. Their countries of origin were Portugal (n¼2), Poland (n¼3), United States
of America (n¼2), Scotland (n¼1), Germany (n¼1), and Italy (n¼1).
4. Photographic documentation and sampling process:
From 2010 to 2014, the lead researcher visited and carefully observed 13 parks and
gardens located in Europe and North America (see Table 2). It is important to stress that
the framework described here is prepared for evaluating the landscape settings found in
8Environment and Planning B: Planning and Design 0(0)
different parks and not the park in its entirety. Therefore, the results should not be
interpreted park-wise and the selection of these parks over others should not lead to any
conclusions.
Site visits and photographic documentation were conducted under specific conditions in
order to ensure the least amount of variability across some key variables related to ‘‘peak
season’’ for vegetation, time of day, and weather. The peak point of vegetation in the park
included peak leafing and flowering periods. This period was considered the most
advantageous for a park visit. Concerning the weather, all photos were taken during
sunny days with a clear sky. This ensured that, in the case of open views, one could
observe a blue sky and, in the case of canopied views, one could see all factors related to
light, shade, and colors. It was also important to select views with no distracting elements,
such as people, animals moving along the vista, falling leaves, vehicles, and so forth (Kaplan
and Kaplan, 1989).
The output of the park visits was a database of photographs out of which only 40 were
selected for the evaluation that was conducted by the participant experts (using the
Questionnaire). Photographs with repeating content were eliminated. In order to cover all
the listed elements of contemplative landscapes and all composition types, at least four
instances per type were kept. Dominant and iconic park vistas were also considered
during the selection process.
5. Online collection of responses:
Based on the eight key categories of the CLM (Landscape Layers, Landform, Vegetation,
Color and Light, Compatibility, Archetypal Elements, Artistic Expression, Character of Peace
and Silence), eight questions were formulated with a 1–6 point Likert rating scale.
These eight questions (# 1–8) constituted the initial part of the Questionnaire.
The following question (# 9) concerned the composition type of the setting, and the last
question (# 10) was a validating question, in which experts were asked to rate the
contemplativeness of the setting according to their experience and knowledge. We used
the Google-Forms tool, which allowed experts to evaluate the selected landscapes at their
own individual pace.
Table 2. Parks and gardens included in the research with the location and designer.
Name of the park Location Designer
1 Parque da Cidade Porto (Portugal) Sido
´nio Pardal
2 Serralves Garden Porto (Portugal) Jacques Greber
3 Parque Andre Citroe
¨n Paris (France) G. Cle
´ment, P. Berger, J.F. Jodry, J.P. Viguier
4 Parchi di Nervi Genoa (Italy) G. Rovelli or M. and G. Roda
5 Central Park New York (NY, USA) Frederic Law Olmsted, Calvert Vaux
6 South Pointe Park Miami (FL, USA) Hargreaves Associates
7 Washington Cathedral Washington DC (USA) Frederick Law Olmsted Jr
8 Chattanooga City Park Chattanooga (TN, USA) Hargreaves Associates
9 Prospect Park New York (NY, USA) Frederic Law Olmsted, Calvert Vaux
10 The National 9/11 Memorial New York (NY, USA) Peter Walker
11 Naumkeag Estate Stockbridge (MA, USA) Fletcher Steele
12 Balboa Park San Diego (CA, USA) Kate Sessions, among others
13 Salk Institute La Jolla (CA, USA) Louis Kahn
Olszewska et al. 9
Results
Statistical measures on the quality of the questionnaire
The quality of the questionnaire was measured in terms of (1) reliability and (2) validity,
which are the most common statistical measures in psychometry. Both reliability and validity
measures have shown satisfactory results, as described below.
Reliability. The reliability of the questionnaire was analyzed through three different methods: (a)
by computing the internal consistency of the first part (questions 1–8), (b) by measuring the
split-half correlation coefficient, and (c) by computing inter-rater reliability (ICC). The
estimated internal consistency of the questionnaire was performed using the Cronbach’s
alpha measure. The standardized Cronbach’s alpha for questions 1–8 was .798, which
signifies good internal consistency. Other outputs of the computation of the Cronbach’s
Alpha internal consistency coefficient are the Item-Total Correlation and the variations of
Alpha if deleted a certain item.The Item-Total Correlation is a correlation of the responses to
each question with the total score of the questionnaire. The measure of Alpha if Deleted
represents the estimated value of Cronbach’s Alpha after deleting a particular item from the
Questionnaire. According to the results of Cronbach’s Alpha computation, questions 1–6, and
8 are substantially informative and contribute to the total score of the Questionnaire (Table 3).
Question 7 appeared not to contribute to the total score. Its correlation with the total
score is close to zero and the overall internal consistency of the survey would be considerably
higher (.817) if this item were absent. For this reason, question 7 was excluded from the final
version of the Questionnaire.
In addition to Cronbach’s alpha, a Guttman split-half test was conducted to examine
reliability, and the obtained result of r
SHG
¼.854 suggests that the questionnaire is highly
reliable.
The final method of reliability analysis performed was the ICC test to measure the
homogeneity between experts when rating a particular landscape. For continuous data,
even if measured through interval or ordinal scales, consensus is measured by inter-class
correlation (ICC).
The inter-rater agreement expressed by the Fvalue and the values of ICC for questions
1–8, as well as for question 10, are presented below. For a single rater ICC ¼.305 (meaning
how accurate ratings would be if a single rater was used), while for the mean of the 10 raters
ICC ¼.814 (F¼6.71, p<.01). ICC values above .80 indicate almost perfect agreement
Table 3. Item-total correlation and Cronbach’s alpha analyses (n ¼40).
Question # Item-total correlation Alpha if deleted
1 .710 .663
2 .552 .705
3 .431 .728
4 .447 .734
5 .831 .679
6 .502 .718
7 .002 .817
8 .545 .715
Notes: Standardized alpha ¼.798; Average inter-item correlation ¼.352.
10 Environment and Planning B: Planning and Design 0(0)
(values above .30 are fair, above .50 represent moderate agreement, and above .70 indicate
strong agreement). The same measures for each item are presented in Table 4.
It is evident that the evaluation of certain characteristics of the landscapes, namely the
ones that are represented in questions 4–6 and 8, become significantly more accurate if a
group of experts appraise such characteristics.
Also, question 7 performed well in the inter-rater agreement test, which means that the
responses of experts concerning the Artistic Expression of the evaluated landscapes were
similar. Therefore, it can be assumed that question 7 was formulated well and understood
correctly by the experts even if Artistic Expression in the landscape design does not have
much to do with the contemplative value of spaces, as the reliability tests have shown,
leading us to exclude Artistic Expression from the final version of the Questionnaire (CLQ).
Fleiss kappa computation (Fleiss, 1971) was used in order to compute the reliability of
question 9. This method assesses the inter-rater agreement between multiple raters when
responses represent categories. question 9 is about the type of landscape composition and
consists of five options (canopied, enclosed, feature, focal, and panoramic), among which
experts were supposed to choose only one (the dominant type). The Fleiss Kappa score was
.304, which signifies fair inter-rater agreement (values of Kappa below .21 are considered to
be of slight or poor agreement).
Validity. The validity of questionnaires in general is not easy to determine in most situations.
In each case, researchers have to find a method of assessing how accurately the data
provided by responses correspond to the real world. (Amer. Educational Research Assn.,
1999)
In the case of our Questionnaire, the validity of the eight items of Part A was tested by
correlating the mean values across experts per image for questions 1–8, with the mean values
for the question ‘‘How contemplative is this landscape for you?’’ (question 10), which served
as a validity question. The Pearson Product-Moment Correlation Coefficient was used for
estimating the validity. This test measures the linear association between two sets of values.
There was a correlation coefficient of r¼.772 (p<.001),which indicates a statistically
significant and strong positive correlation between combined responses to questions 1–8
and how contemplative landscapes are from the experts’ judgment (Figure 2). Testing the
validity was also performed by measuring the statistical significance of the correlations
between the mean response scores of each question and question 10 (Table 5).
Table 4. Inter-class correlation for each item.
Question #
ICC
(single rater)
ICC (mean of
10 raters) Fp
1 .417 .877 9.43 <.01
2 .463 .896 11.9 <.01
3 .444 .889 9.65 <.01
4 .068 .423 1.78 <.01
5 .130 .599 2.74 <.01
6 .094 .510 2.37 <.01
7 .341 .838 8.16 <.01
8 .152 .642 3.08 <.01
Mean .305 .814 6.71 <.01
Olszewska et al. 11
In an item-to-item analysis, there were very strong positive correlation coefficients
between questions 5 and 8, and the validity question; strong positive correlations between
questions 1, 3, 4, and 6, and the validity question. Question 2 showed a moderate correlation
(positive) with the validity question, while question 7 has a negligible relation to the validity
question. In fact, all correlations were statistically significant (all p<.05), with the exception
of question 7 and the validity question. Therefore, the decision about deleting question 7
from the final version of the CLQ was reinforced.
In Table 6, there is a synthesis of the results concerning the reliability and validity
analyses that were performed on the questionnaire and its questions regarding the
contemplative characteristics of landscapes.
Ratings of the landscapes
A default output of this method is a ranking of tested photographs. It can be used for
landscape research, together with the Questionnaire or without it. In order to use any of
Figure 2. Correlation between two sets of data: mean of responses for questions 1–8, and question 10
(validity question).
Table 5. Correlations between questions 1–8
and question 10.
Question # Item 10 (val) p
1 .614 <.001
2 .339 .032
3 .444 .004
4 .447 .004
5 .837 <.001
6 .559 <.001
7 .119 .463
8 .901 <.001
Mean .772 <.001
Notes: Marked correlations are significant at p <.05.
12 Environment and Planning B: Planning and Design 0(0)
the 40 landscapes in studies where contemplation is a variable of interest, researchers need a
classification system or criteria to select subsets of the most contemplative and the least
contemplative landscapes out of the ranking.
For example, considering that the mean score for the 40 landscapes was 4.05 (on the
6-point Likert scale) with a standard deviation of 0.562 (Max ¼5.12; Min ¼2.82), each
landscape with a score of half standard deviation above the average (i.e. scoring higher
than 4.33 on contemplative features) may be selected for the most contemplative set. On
the other hand, each photo with a score of half standard deviation below the average
(i.e. scoring lower than 3.77) may be included in the least contemplative set. The criteria
may be easily adjusted in order to select a lower number of landscapes for each subset,
namely by increasing the standard deviation around the mean score. This procedure may be
particularly useful when only extreme contemplative and noncontemplative landscapes are
desired. To that purpose, Figure 3 presents the three photos that scored the highest in
contemplativeness and the three that scored the lowest.
Experts classified the majority (43%) of the most contemplative landscape settings as
panoramic, 29% as focal, 14% as feature, and 14% as enclosed composition type. None
were classified as a canopied landscape (see Figure 4). Experts classified the majority
(50%) of the least contemplative landscape settings as enclosed, 30% as canopied, 10% as
afeature, and 10% as a focal composition type. None were classified as a panoramic
landscape (see Figure 4).
Discussion
Through this study, a tool has been developed to evaluate landscape settings in terms of their
contemplativeness, which in its final version is the CLQ.
Through the developed framework, the question of what makes landscapes contemplative
has been answered and empirically explored. The contemplativeness of the landscape is
determined by seven key elements, which can be identified on photographs of landscape
settings and evaluated using the CLQ (see Figure 5 in Appendix 1).
In the study, eight key components of the landscape setting were tested in terms of
reliability and validity. The reliability measures showed how much each of the eight items
contributed to the overall score. Items 1, 5, and 8 (Landscape Layers,Compatibility, and
Character of Peace and Silence, respectively) provided the highest contribution to the
contemplative character of the space (Table 3). Also, items 2, 3, 4, 6 (Landform,
Vegetation, Light and Color, Archetypal Elements, respectively) demonstrated importance.
On the other hand, the contribution of question 7 (Artistic expression) to the overall score
was negligible, diminishing the overall score of the Questionnaire in terms of reliability.
Table 6. Summary of the reliability and validity indexes of the questions 1–8 of the questionnaire on
contemplation features and their meaning.
Statistical
index Name of the test Measure/result Meaning
Reliability Cronbach’s Alpha
(internal consistency)
Alpha ¼.798
(if Item 7 deleted, alpha ¼.817)
Acceptable
(Good, if Item 7 deleted)
Gutmanns’s split-half Alpha ¼.854 Very good
inter-rater agreement ICC (10 raters) ¼.81 Almost perfect agreement
Validity Pearson’s correlation
coefficient
r¼.772 for the mean score Strong positive correlation
Olszewska et al. 13
It may be that artistic expression has less to do with the overall scene perception in general
and the other aspects of the scene (e.g. character of peace and silence) influence the overall
setting perception more significantly. Considering the Cronbach’s alpha value of question 7
(.001), it seems that the level of its importance is so low to the construct of contemplativeness
Figure 3. Three photos that scored the highest and three photos that scored the lowest in
contemplativeness. (a) Most contemplative scenes and (b) least contemplative scenes.
Figure 4. Distribution of (a) most contemplative (CL) and (b) least contemplative (NCL) landscape
pictures across composition types.
14 Environment and Planning B: Planning and Design 0(0)
that, if preserved in the Questionnaire, it would deteriorate the overall quality of future
studies.
The finding above may seem controversial, because the contemplation of nature has often
been compared to the contemplation of art. Also, art is typically considered to be an
important factor in improving the quality of public space. However, in light of the
presented study, it may be more reasonable to distinguish natural and artistic
contemplation as two separate concepts. In any case this issue needs further research.
The study shows that, in terms of landscape composition, panoramic and focal are the
most contemplative, and enclosed and canopied are the least contemplative landscape
composition types (Figure 4). This suggests that open landscapes with long-distance views
are the most contemplative, no matter if it is a wide-open vista or an axis leading to one
distant point. Moreover, Landscape Layers (question 1) show the highest contribution to the
overall contemplativeness of the scene. In addition, enclosed and canopied landscapes are
designated as much less contemplative, which suggests the contemplative experience largely
depends on how far one can see into the landscape.
Some existing research seems to confirm this statement. For example, the previously
mentioned study about the comfort of the long-distance view (Skalski, 2005), and the
Savannah hypothesis, which proposes that people prefer open spaces or sparsely wooded
areas rather than dense, complex environments (such as forests or mountains), because of
their anatomical and psychological construction (Kaplan and Kaplan, 1989; Lidwell et al.,
2003), would appear to support these findings.
The results of this research also seem to correspond to the findings of environmental
psychology regarding studies on landscape preference. According to such studies, people
prefer landscape settings with long-distance views framed with dense vegetation, even
though they found the desert and prairie landscapes less interesting (Kaplan et al., 1998).
However, the canopied or enclosed composition type, which is often a preferred landscape,
was rated low in contemplation. This may suggest that the contemplative experience of the
landscape needs more open sky and distant views in some cases or at least a far-away focus
within an enclosed landscape. The difference between landscape preference and
contemplative landscape is probably connected with the research scope and balances
between public versus expert-based surveys.
One limitation of the presented framework is item 9 (composition type), which is based on
the already existing landscape composition classification developed by Richard Smardon
et al. (1986). This item showed fair inter-rater agreement, which is not considered
satisfactory in terms of reliability. It means that experts were not consistent about the
composition types of the landscape. This is why including or rejecting this item should be
carefully considered in future studies. It would not be recommended to use item 9 for
evaluation of a small number of landscape settings. Some general conclusions were based
on this item (Figure 4), mainly due to the large number of landscape settings and to illustrate
some general tendencies.
Conclusions
Creating landscapes for contemplative purposes has been present in the space-design
professions for centuries, but a lack of common understanding and an operational definition
has limited contemplativeness to the artistic and spiritual, excluding the technical and practical.
The presented study has shown that it is possible to operationalize a concept such as
contemplativeness of a space and to create a reliable and valid tool (CLQ) for evaluating the
contemplativeness of designed outdoor spaces (mainly, urban parks and gardens).
Olszewska et al. 15
The contemplativeness of the space depends on seven key components, identified as:
Layers of the Landscape, Landform, Vegetation, Light and Color, Compatibility, Archetypal
Elements, and Character of Peace and Silence. The most contemplative landscapes can be
then defined as ones with a high level of compatibility, characterized by adjacent scenery and
all elements and views of the landscape worked out in terms of scale, balance, and harmony.
Also, those with long-distance views, where the observer is able to see the fore, middle, and
background, with a smooth landform and vegetation seemingly native, but maintained and
organized, where all elements of composition are worked out in terms of scale and inter-
relations, where colors are natural, and not too contrasting, but one can observe light and
shade movements. The most contemplative landscapes contain archetypal elements, such as
a water mirror or an old single oak tree, and invite rest and relaxation through their
character of peace and silence.
The findings described above confirm that contemplative landscapes are a measurable
reality.
The importance of defining contemplative landscapes is directly connected to the
improvement of the quality of the designed urban parks and gardens, which leads to
improvement in the quality of life of city inhabitants as well as benefits in terms of their
mental health and well-being.
This study is relevant for design professionals, researchers, and anyone interested in
improving the quality of life in urbanized areas. Landscape architects and urban planners
can use the patterns illustrated in the CLM to plan new parks and gardens for improving the
quality of life and well-being in cities, contributing to mental health through attention
restoration and stress reduction. The CLQ can also be used for evaluation of specific,
already existing landscape scenes, i.e. to evaluate their beneficial potential, reprogram,
protect, or compare with others. The findings of this study can be used for reinforcing the
design process with evidence, in the continuous endeavor of improving the quality of the
built environment and the well-being of urban inhabitants.
For future research, this study presents the opportunity to engage in a more profound
investigation that links the effects of the designed environment with the well-being of people.
One suggested focus could be to continue the investigation of this research team on the brain
responses to different kinds of landscape settings. Building bridges between seemingly distant
areas of the arts and sciences is still in its infancy, but is essential for better understanding the
interrelations between man and his living environment.
Acknowledgements
The research presented in this paper was conducted with contribution of the team of Laboratory of
Neuro-psychophysiology of the University of Porto.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or
publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or
publication of this article: The presented research was funded by the Portuguese National Foundation
for Science and Technology (FCT) with the individual doctoral grant for the lead author (ref# SFRH/
BD/77141/2011).
16 Environment and Planning B: Planning and Design 0(0)
References
AERA & NCME, American Educational Research Association, Psychological Association, &
National Council on Measurement in Education, Washington, DC (1999) Standards for
Educational and Psychological Testing. Washington, DC: AERA Publications.
Bell S (1998) Landscape Pattern, Perception and Process. London and New York: E & FN.
Burney J, Calle I, Jia J, et al. (2007) Boston Harbor Islands National Park Area Scenic Analysis and
Assessment: A Pilot Study. Landscape Architecture and Regional Planning. Amherst, MA, USA:
University of Massachusetts – Amherst. Available at: http://scholarworks.umass.edu/larp_grad_
research/35/ (accessed 20 March 2014).
Chiesura A (2004) The role of urban parks for the sustainable city. Landscape and urban planning 68(1):
129–138.
Delbecq AL, Van de Ven AH and Gustafson DH (1975) Group Techniques for Program Planning.
Glenview, IL: Scott, Foresman and Co.
Eberhard JP (2009) Brain Landscape, The Coexistence of Neuroscience and Architecture. Oxford:
Oxford University Press.
Fleiss JL (1971) Measuring nominal scale agreement among many raters. Psychological Bulletin 76(5):
378–382.
Harris CW and Dines NT (1988) Time-Saver Standards for Landscape Architecture: Design and
Construction Data. Columbus, OH: McGraw-Hill.
Hermann H (2005) On the transcendent in landscapes of contemplation. In: Krinke R (ed.)
Contemporary Landscapes of Contemplation. London and New York: Routledge, pp. 36–73.
Herzog TR and Barnes GJ (1999) Tranquility and preference revisited. Journal of Environmental
Psychology 19(2): 171–181.
Herzog TR, Maguire P and Nebel MR (2003) Assessing the restorative components of environments.
Journal of Environmental Psychology 23(2): 159–170.
Hou R and Renzhi (2015) From Beijing to Washington – A contemplation in the concept of
municipal planning. In: Symposium on Chinese Historical Geography. Berlin Heidelberg: Springer,
pp. 61–82.
Hsu CC and Sandford BA (2007) The Delphi technique: Making sense of consensus. Practical
Assessment, Research and Evaluation 12(10): 1–8.
Jung CG (1955) The Archetypes and the Collective Unconscious. London: Routledge.
Jung CG (1964) Man and His Symbols. New York: Doubleday.
Kaplan S (1995) The restorative benefits of nature: Toward an integrative framework. Journal of
Environmental Psychology 15(3): 169–182.
Kaplan R and Kaplan S (1989) The Experience of Nature: A Psychological Perspective. New York:
Cambridge University Press.
Kaplan R, Kaplan S and Ryan R (1998) With People in Mind: Design and Management of Everyday
Nature. Washington, DC: Island Press.
Kaplan S, Kaplan R and Wendt JS (1972) Rated preference and complexity for natural and urban
visual material. Perception & Psychophysics 12(4): 3542–356.
Krinke R (2005) Contemplative landscapes, restorative landscapes. In: Krinke R (ed.) Contemporary
Landscapes of Contemplation. London and New York: Routledge, pp. 107–139.
Krippendorff K (1995) Redesigning Design; an Invitation to a Responsible Future. Helsinki, Finland:
University of Pennsylvania Scholarly Commons.
Lidwell W, Butler J and Holden K (2003) Universal Principles of Design: 100 Ways to Enhance
Usability, Influence Perception, Increase Appeal, Make Better Design Decisions, and Teach
Through Design. New York: Rockport.
MArkwell KW (2000) Photo-documentation and analyses as research strategies in human geography.
Australian Geographical Studies 38(1): 91–98.
Neufert E and Neufert P (2012) Architects’ Data. Hoboken, NJ: Wiley-Blackwell.
Norberg-Schulz C (1976) The phenomenon of place. The Urban Design Reader 3: 125–137.
Olszewska et al. 17
Olszewska A (2015) Contemplative values of urban parks and gardens: Applying neuroscience to
landscape architecture. PhD Thesis, University of Porto, Porto, Portugal.
Pardal S (2006) Porto City Park, Idea and Landscape. Lisbon: GAPTEC.
Skalski J (2005) Comfortof long-distance perceiving and a landscapeof river valley in towns situated on the
plains. Teka Komisji Architektury, Urbanistyki i Studiœw Krajobrazowych. (OL PAN, Krakow) 1: 44.
Smardon RC, Palmer JF and Felleman JP (1986) Foundations for Visual Project Analysis. New York:
Wiley.
Stigsdotter U (2005) Landscape architecture and health, evidence-based health-promoting design and
planning. PhD thesis, Swedish University of Agricultural Sciences, Alnarp.
Tatler BW (2007) The central fixation bias in scene viewing: Selecting an optimal viewing position
independently of motor biases and image feature distributions. Journal of Vision 7(14): 4.
Treib M (2005) Attending. In: Krinke R (ed.) Contemporary Landscapes of Contemplation. London and
New York: Routledge, pp. 13–36.
Tuan YF (1974) Topophilia: A Study of Environmental Perception, Attitudes, and Values. Englewood
Cliffs New Jersey: Prentice-Hall.
Ulrich RS and Parsons R (1992) Influences of passive experiences with plants on individual well being
and health. In: Relf D (ed.) The Role of Horticulture in Human Well Being and Social Development.
Portland, OR: Timber Press, pp. 99–100.
U.S.D.A., Forest Service. (1973) National Forest Landscape Management, Volume 1. Washington, DC:
U.S. Government Printing Office.
Windhager S (2009) Sustainable Sites Initiative. Austin, Texas, USA: University of Texas at Austin,
School of Architecture.
World Health Organization (2005) Mental Health: Facing the challenges, building solutions. Report
from the WHO European Ministerial Conference, Helsinki.
Zelanski P and Fisher MP (1996) Design Principles & Problems. New York: Brace College.
Agnieszka Anna Olszewska is a PhD in Landscape Architecture and Urban Ecology, recently
defended her thesis titled: ‘‘Contemplative Values of Urban Parks and Gardens: Applying
Neuroscience to Landscape Architecture’’ at the University of Porto. She is interested in the
evidence-based design and the impact of the design of the public spaces on the mental health
and well-being of people.
Paulo Farinha Marques, is an Auxiliary Professor at the University of Porto where he
currently develops teaching and research in the landscape architecture field – landscape
design, landscape planting, ecological park design, design and restoration of natural and
manmade habitats and biodiversity promotion in historic landscapes.
Robert L Ryan is a Professor at the Department of Landscape Architecture and Regional
Planning at University of Massachusetts – Amherst. His research addresses the question
‘‘What motivates people to become engaged in sustainable landscape design, planning, and
management practices that benefit the environment, and how does that affect their attitudes
and behaviors in the landscape?’’. His current research focuses on visual resource
management, greenway and green infrastructure planning, and sustainable site design.
Fernando Barbosa is a Professor at the at the Faculty of Psychology and Educational
Sciences, University of Porto. Coordinator of the Laboratory of Neuro-psychophysiology.
Interested in psychobiological research on antisocial behavior in general and psychopathy in
particular.
18 Environment and Planning B: Planning and Design 0(0)
Appendix 1
Figure 5. Contemplative Landscape Questionnaire (CLQ)—final version.
Olszewska et al. 19
