Architectural Lessons From Environmental Psychology : The Case of Biophilic Architecture

Review of General Psychology
Copyright 2007 by the American Psychological Association
2007, Vol. 11, No. 4, 305–328
1089-2680/07/$12.00
DOI: 10.1037/1089-2680.11.4.305
Architectural Lessons From Environmental Psychology:
The Case of Biophilic Architecture
Yannick Joye
Free University of Brussels
A review of ?ndings from the ?eld of environmental psychology shows that humans are
aesthetically attracted to natural contents and to particular landscape con?gurations.
These features are also found to have positive effects on human functioning and can
reduce stress. However, opportunities for contact with these elements are reduced in
modern urban life. It is argued how this evolution can have subtle but nontrivial adverse
effects on psychological and physiological well-being. These can be countered by
integrating key features of natural contents and structural landscape features in the built
environment. Several practical proposals are discussed, ranging from literal imitations
of natural objects (such as plants) to the use of nature’s fractal geometry in an
architectural context.
Keywords: biophilia, environmental psychology, fractal architecture, biophilic archi-
tecture
Natural objects, shapes, and processes have
tions in architecture are valuable for human
often acted as a source of inspiration throughout
emotional and cognitive functioning. However,
the history of architecture. Perhaps the most
the exact way in which this conclusion is
obvious example of this inspiration is ornament,
reached differs in an important respect from the
which often contains representations that are
narratives and arguments proposed in theories
closely similar to, or reminiscent of, the animal
of organic and biomorphic architecture, which
and plant world. Besides such literal imitations,
are often more philosophical (Lynn, 1998) or
some architects, notably Anton?´ Gaud?´, drew
even pseudophilosophical (e.g., Steiner, 1999)
lessons from the structural forces governing nat-
in nature. In contrast, the argument for nature-
ural structures, resulting in ef?cient and eco-
based forms in architecture in the current study
nomically built architecture (e.g., Sweeney &
is mainly based on empirical ?ndings from di-
Sert, 1960). Today, there seems to be a renewed
verse psychological subdisciplines. In particu-
interest in the relation between nature and ar-
lar, the article starts with a concise review of
chitecture, especially in zoomorphic or biomor-
empirical ?ndings from the ?eld of environ-
phic architecture (e.g., Feuerstein, 2002). More
mental psychology and aesthetics. This survey
speci?cally, such architecture makes use of dig-
reveals that humans have an emotional relation
ital design software, which allows one to easily
with natural elements and shows that contact
recreate the curvy shapes and geometry that are
with natural form is in a sense good for human
characteristic of natural entities (Lynn, 1999).
psychological and physiological functioning. It
This article af?rms the importance of natural
is argued that, by architecturally mimicking nat-
form as a perennial source of inspiration for
ural forms and structural organizations of natu-
architecture. In fact, the main conclusion of this
ral settings, these bene?cial effects can be
study is that nature-based forms and organiza-
tapped in a built context.
The Psychoevolutionary Framework
Yannick Joye, Business Economics and Strategic Policy
Different psychological sub?elds study the
(BEDR), Free University of Brussels, Belgium.
human relation with nature. For example, evo-
Correspondence concerning this article should be ad-
lutionary psychologists argue for the existence
dressed to Yannick Joye, Business Economics and Strategic
of cognitive modules that are specialized in
Policy, Pleinlaan 2, B-1050 Brussels, Belginm. E-mail:
yannick.joye@telenet.be or yannick.joye@vub.ac.be
conceptual and perceptual knowledge about
305

---

306
JOYE
natural entities. Such cognitive devices are
risk (e.g., turbulent water, a predator), this trig-
claimed to have evolved to handle the survival-
gered negatively toned affective reactions (e.g.,
related challenges and opportunities that were
dislike), ultimately leading to avoidance behav-
present in the natural settings in which human
ior. On the other hand, if a setting offered good
ancestors lived (e.g., ?nding food; Mithen,
opportunities for survival and reproduction, this
1996; Pinker, 1994). Scott Atran (1995) argues
would have caused liking reactions, leading to
how cross-cultural similarities in human folkbi-
explorative behavior.
ologies support the existence of such a system.
In agreement with Ulrich’s (1983) model,
In particular, Western and non-Western individ-
empirical evidence shows that environments
uals seem to classify nature in similar ways and
(e.g., urban vs. natural) are processed according
consistently ascribe essences to the taxonomic
to their affective valence. Moreover, this pro-
types of folkbiologies. Atran’s view is consis-
cess seems to occur very rapidly, which sup-
tent with the literature on category-speci?c def-
ports the immediate and automatic character of
icits, in which people are reported to have de-
these affective responses (Hietanen & Korpela,
?cient perceptual and conceptual knowledge
2004; Korpela Klemettila¨, & Hietanen, 2002).
about the category of living things. One of the
According to the psychoevolutionary frame-
interpretations of the causes of such de?cits is
work, survival chances further increased if these
that, under evolutionary pressures, speci?c neu-
emotional reactions had an inherited compo-
ral areas have become specialized in informa-
nent: No precious time and energy had to be
tion about living entities (e.g., Caramazza &
spent learning what kinds of environments were
Shelton, 1998).
either bene?cial or harmful (S. Kaplan, 1987,
Perhaps the psychological ?eld that has most
1988; Ulrich, 1983). With regard to the neural
profoundly studied the human (affective) psy-
origin of these affective states, some researchers
chological relation with natural entities is envi-
attribute an important role to subcortical areas,
ronmental psychology. This research area draws
especially the amygdala. Because these struc-
on numerous empirical studies and is, therefore,
tures are also involved in modulating stress-
less speculative than the just-mentioned modu-
related hormones, it provides an explanation of
larity thesis. One of the central issues of envi-
why certain types of settings have a different in-
ronmental psychology is how different types of
?uence on autonomic stress responses (Parsons,
settings can trigger different affective states in
1991; see also Joye, 2007, for an in-depth discus-
individuals (e.g., liking or preference reactions).
sion of this issue).
Two important proposals have been advanced
with regard to the speci?c process underlying
Aesthetic Preference and Structural
these emotional states. In the preference matrix,
Landscape Features
developed by Stephen and Rachel Kaplan, the
occurrence of such states is to a large extent the
What is the character of the settings or ele-
result of cognitively assessing whether certain
ments that can trigger such immediate affective
informational features are present in a setting
states? The literature states that, on the one
(R. Kaplan & Kaplan, 1989). This contrasts
hand, these reactions can be provoked by some
with a central tenet of Roger Ulrich’s (1983)
typical structural landscape features. Although
psychoevolutionary framework. In this model,
coming from a different research ?eld, geogra-
which will form the backbone of the current
pher Jay Appleton was one of the ?rst to pro-
study, affective responses toward environmen-
pose a model addressing this issue (Appleton,
tal settings are not mediated by cognition but
1975). According to Appleton’s prospect–
stem from a rapid, automatic, and unconscious
refuge theory, human beings’ preference for
process by which environments are immedi-
landscapes correlates with two environmental
ately liked or disliked. These fast affective re-
qualities: prospect and refuge. The notion of
actions are claimed to be rooted in human evo-
prospect refers to settings or landscape elements
lutionary history and are essentially adaptive:
that facilitate obtaining information about the
They motivated the organism to quickly under-
environment. A typical example is a hill, which
take actions that contributed to its well-being
aids to visually access and inspect the surround-
and survival. For example, if early humans
ing area. On the other hand, refuge points to
came across a setting containing an important
settings that can provide shelter and protection.

---

ARCHITECTURAL LESSONS FROM ENVIRONMENTAL PSYCHOLOGY
307
An evident example is a cave, which can protect
liked because they contributed to the survival
against predators and weather conditions.
and reproduction of early humans. Flowers, for
Ulrich’s psychoevolutionary framework lists
example, signaled the presence of food sources
some other visual cues that are associated with
and were cues for future foraging sites. They
immediate positive affective reactions: com-
also helped in differentiating between different
plexity, gross structural properties (e.g., pat-
vegetation types, because plants that are not
terns), depth properties, ground surface and
blooming often look quite similar (Orians &
texture, absence of threats, and de?ected vista
Heerwagen, 1992). Trees protected against sun
(Ulrich, 1983). The predictors in Rachel and
and rain and offered early humans prospects on
Stephen Kaplan’s preference matrix overlap to
the surrounding landscape and retreats from
a certain extent with the variables listed by
predators (Appleton, 1975; Orians & Heerwa-
Ulrich. The Kaplans’ model (R. Kaplan &
gen, 1992; Summit & Sommer, 1999).
Kaplan, 1989; S. Kaplan, 1987, 1988) describes
These bene?ts can explain why vegetative
two types of postures toward the environment.
elements and settings containing vegetation still
An individual can be actively involved in an
cause aesthetic or liking reactions. Different
environment: One can, for example, explore the
empirical studies show that individuals consis-
setting. Alternatively, an individual can try to
tently prefer natural, vegetated landscapes over
understand the environment. The Kaplans argue
urban settings without vegetation. When urban
that these two attitudes are associated with four
environments are mutually compared, highest
structural landscape properties, each of which
preference is associated with urban settings con-
correlates with positive aesthetic evaluations
taining some vegetation, especially trees, or a wa-
and positively in?uences landscape selections.
ter feature (Smardon, 1988; Thayer & Atwood,
The structural properties that facilitate involve-
1978; Ulrich, 1986). Such phytophilia is also clear
ment in the environment are complexity and
from the observation that nonnatural environ-
mystery. Stephen Kaplan de?nes complexity as
ments (e.g., home and working interiors) often
a measure for “how much is ‘going on’ in a
contain actual vegetative elements or (decorative)
particular scene, how much there is to look at”
references to natural content (Eibl-Eibesfeldt,
(1988, p. 48). Mystery refers to settings whose
1989; Heerwagen & Orians, 1986). Although em-
layout suggests that more information can be
pirical research on preferential reactions toward
acquired if the scene is penetrated more deeply.
?owers is scarce, some studies show that these
An example of a mysterious landscape quality is
elements are indeed associated with positive aes-
a de?ected vista, such as a winding trail. Struc-
thetic reactions (e.g., Haviland-Jones, Rosario,
tural properties that facilitate understanding the
Wilson, & McGuire, 2005; Todorova, Asakawa,
environment are coherence and legibility. Co-
& Aikoh, 2004).
herence refers to features that contribute to the
Still, a possible critique of preferential re-
organization, understanding, and structuring of
search into greenery is that when vegetated
the landscape image, such as symmetries, re-
landscapes and nonvegetated (urban) architec-
peating elements, and unifying textures. Finally,
tural settings are mutually compared, the latter
legibility refers to landscape qualities that help
most often involve representations of quite
to predict and maintain orientation in the land-
modern buildings or at least buildings that are
scape as one further explores it. Think, for ex-
not very rich in form. However, it can be
ample, of a prominent rock functioning as an
pointed out that nature is often characterized by
orientation point.
a typical sort of geometry, or fractal geometry
(see later discussion). This type of geometry
The Aesthetic Appeal of Natural Contents
often does not apply to modern buildings or
modern urban settings. A possibility that needs
In addition to the previous structural land-
to be entertained is that the preference for veg-
scape features, the ?eld of environmental psy-
etated scenes is not due to the fact that it is a
chology also studies the natural contents that
natural setting but, instead, that it must (to a
contribute to the aesthetic qualities of settings:
certain extent) be drawn back to the underlying
namely (calm) water features and vegetative
geometric features of the scene. It would, there-
elements. The explanatory framework, again, is
fore, be interesting to compare natural settings
essentially evolutionary. These elements are
with buildings or urban scenery that emulate a

---

308
JOYE
more natural geometry, such as Gothic cathe-
Within the ?eld of landscape aesthetics, the
drals. (Note that this critique does not under-
savanna hypothesis is often taken for granted
mine our plea for nature-based, or biophilic,
and has remained mostly undisputed. Neverthe-
architecture but instead only strengthens it.)
less, we ?nd it troubling that almost no attention
is paid to discussions in the ?eld of paleoan-
The Savanna Hypothesis
thropology. For instance, Wilson (1993) argues
how our preferences for nature, savannas in
In habitat theory, savannas are claimed to
particular, are remnants of paleohominid and
be the settings in which early humans spent a
early Homo evolution in this type of biome, a
substantial part of their evolutionary history,
view shared by many in the ?eld of habitat
and these seem to display an ideal mix of the
theory and landscape aesthetics. Yet there is no
previously discussed structural landscape fea-
consensus on the claim that the savanna is the
tures and natural contents (Van den Berg,
unique environment of evolutionary adapted-
2004). This type of biome can be broadly
ness. In his review, Potts (1998) sketches a
described as low to intermediately complex
more complex view that is supported by scien-
settings, having a relatively even and grassy
ti?c environmental analyses. It evidences that,
ground surface dotted with scattered trees or
during the evolution of early hominids, there
tree groups. Savannas contain a high degree
was quite some variation in the environments
of biomass and meat, and these are relatively
that were inhabited, ranging from forests to
easily accessible for terrestrial beings (as op-
savannas to open-canopy woodlands. Still, it
posed to, e.g., tropical forests). Furthermore,
could be countered that the truth value of the
the openness of savannas facilitates detecting
savanna hypothesis does not have any bearing
predators and game and is conducive of
on the ?nding that humans adapt to and display
movement and a nomadic lifestyle (Orians,
positive affective af?liations with natural envi-
1980; Orians & Heerwagen, 1992).
ronments. As Kahn (1999, p. 39) points out,
The aesthetic preference for savanna-type
“The evolutionary account can hold, but the
landscapes has been the subject of a few
savanna hypothesis needs to give way to a
empirical studies. Balling and Falk (1982)
broader account of genetic predispositions to
found that young individuals preferred savan-
inhabited landscapes.”
nas over other biomes without ever being
exposed to the former type of landscape. The
Naturalness and Stress Reduction
researchers hypothesize that these ?ndings
could well point to an innate (aesthetic) pref-
Besides causing liking responses, natural el-
erence for savannas (see also: Synek & Gram-
ements (e.g., vegetation and water features) are
mer, 1998; but see Coss, 2003, for contrasting
also found to contribute to the restoration of
?ndings). Consistent with the savanna hy-
human individuals. Two major interpretations
pothesis, research indicates that people tend
of restorative responses have been proposed.
to prefer tree shapes characteristic of high-
The ?rst, attention restoration theory (ART),
quality savannas: These typically have a low
was developed by the Kaplans (e.g., R. Kaplan
trunk, a broad canopy, and a moderate canopy
& Kaplan, 1989). Essentially, ART interprets
layering (Orians & Heerwagen, 1992; Sommer
restoration as the recovery of directed attention
& Summit, 1995). Inquiries into the evolution
or the ability to focus. This capacity is deployed
of artists’ work (e.g., John Constable) show an
during tasks that require profound concentra-
increase of conspicuous savanna features over
tion, such as proofreading or studying. Natural
time (e.g., opening up views; Heerwagen &
settings have been found to be ideally suited to
Orians, 1993). Furthermore, studies indicate
restore or rest directed attention (e.g., Hartig,
that when arti?cial changes are made to plants
Evans, Jamner, Davis, & Ga¨rling, 2003; Hartig
and trees, these increasingly come to resemble
Mang, & Evans, 1991).
savanna-type vegetation (Heerwagen & Orians,
The second major interpretation of restora-
1993). Finally, areas that are created for recre-
tion is a part of Ulrich’s psychoevolutionary
ational or aesthetic purposes (e.g., parks or golf
framework. In this view, restoration applies to a
terrains) often resemble savannas (Orians,
much broader context than attentional capaci-
1980).
ties (e.g., Parsons, 1991; Ulrich, 1993; Ulrich et

---

ARCHITECTURAL LESSONS FROM ENVIRONMENTAL PSYCHOLOGY
309
al., 1991). More speci?cally, Ulrich under-
The Value of Nature-Based, or Biophilic,
stands restoration as stress reduction, and stress
Architecture
can occur even when directed attention is not
fatigued. Within Ulrich’s model, restorative re-
Although there is solid empirical evidence
sponses are explained by the fact that early
that humans hold positive af?liations with a
humans were often confronted with threatening
speci?c set of landscapes and natural elements,
and demanding situations (e.g., a predator). As
this does not preclude that some natural features
discussed, such confrontations lead to the quick
or occurrences also cause more negative and
onset of negatively toned affective reactions and
even aversive reactions in humans (e.g., Mineka
corresponding adaptive behavior. Ulrich (1993)
& O
¨ hman, 2002; Van den Berg & ter Heijne,
2005). Another issue is that the experimental
notes that the immediate effects of such re-
outcomes are sometimes dif?cult to integrate
sponses are bene?cial for the individual. Yet
into a coherent, overarching conceptual frame-
such reactions also have a certain cost in that
work. For example, what is the relation between
they lead to physiological and psychological
abstract structural landscape features and the
stress (e.g., high blood pressure, feeling de-
preference for water elements? Other issues are
pressed). When the threat has vanished, the
that differences in nature appreciation are often
individual is in need of restoration from the
left in the dark in these discussions (but see Van
stress that has been caused. The bene?ts of such
den Berg, 2004), and it is still a matter of debate
restorative responses are “a shift toward a more
in which sense the genetic component of these
positively toned emotional state, mitigation of
(positive) af?liations should be understood
deleterious effects of physiological mobiliza-
(Cummins & Cummins, 1999). Still, whatever
tion (reduced blood pressure, lower levels of
the outcome of these matters, the general pic-
circulating stress hormones), and the recharging
ture emerging from the previous concise review
of energy expended in the physiological arousal
is that humans have a (partly) hardwired emo-
and behavior” (Ulrich, 1993, p. 99). These re-
tional af?liation with certain classes of natural
storative responses typically occurred in natural
objects. Some researchers have argued about
unthreatening (savanna-like) settings. Such
the affective relation with natural elements and
open, low-risk environments often contained a
landscapes in terms of biophilia (e.g., Kellert &
(calm) water feature and sometimes had a small
Wilson, 1993; Wilson, 1984). Although the the-
?re. Restoration was also facilitated by the avail-
oretical merits of this term have been ques-
ability of resources, which reduced stress related
tioned (Joye, 2007), in the remainder of this
to the uncertainty of ?nding food (Ulrich, 1993).
report the notions biophilic and biophilia will
The stress-reducing effect of nature is still
nevertheless be used as synthetic concepts.
effective today because those individuals who
The occurrence of biophilic responding
could respond restoratively to stressful situa-
stands in sharp contrast with the observation
that there is increasingly less contact with na-
tions survived better. In an often-cited article in
ture in Western technologically oriented societ-
Science, Ulrich (1984) discusses a study of hos-
ies. Wolff, Medin, and Pankratz (1999) found
pital patients who had undergone gall bladder
that such an evolution has nontrivial effects on
surgery and had rooms with views of either a
cultural expressions of nature. In particular,
small tree group or of a brown brick wall. As
they made a historical study of word use in
opposed to patients with the brick wall view,
dictionaries and found that, from the 20th cen-
patients with the tree view had shorter hospital
tury onward, the use of (folk)biological terms
stays, received fewer negative comments from
devolved, and their application lost precision. In
the nurses, required less moderate and strong
contrast, several nonbiological terms evolved
analgesics, and had slightly fewer postoperative
during this period (e.g., books, clothes, furni-
complications. (For further research into the
ture). Apart from being associated with an im-
relation between stress reduction and nature,
poverishing conceptual framework for natural
see, e.g., Custers, 2006; Hartig et al., 2003;
objects, it is also plausible that reduced contact
Parsons, Tassinary, Ulrich, Hebl, & Grossman-
with nature can be accompanied by a reduced
Alexander, 1998; Ulrich, 1981; Ulrich et al.,
knowledge of the rich variety of forms charac-
1991).
teristic of natural entities. A probable artistic or

---

310
JOYE
creative consequence is that the formal curric-
them to other domains or contexts can prove
ulum of artists and architects becomes nar-
problematic and harmful. Today, we can wit-
rower. The reason is that natural form can be
ness how thinking about nature and natural re-
considered as a creative or compositional gram-
sources in terms of things that can be manipu-
mar, which can be used for creating artwork, or,
lated has devastating effects. The upshot is that
as Stephen Kellert put it, “The aesthetics of
this shifts the balance even further toward func-
nature can function as a kind of monumental
tional thinking, because nature is replaced by
design model” (Kellert, 1997, p. 36). The loss of
entities that predominantly require functional
this monumental design model has its architec-
analyses. Probably, this process can be coun-
tural counterpart in modern urban settings,
tered by extensive contact with the natural
which are increasingly governed by euclidean
world, and developing a rich conceptual frame-
geometry and stripped of ornament, patterning,
work about it (e.g., by nature education). This
detailing, and color (Salingaros, 2004). Archi-
could help people realize that functional think-
tectural references to nature can help put an end
ing is not always desirable with regard to na-
to this uniformity. By encouraging architects to
ture. While being more speculative, it can be
integrate natural forms and patterns in their
hypothesized that integrating naturalistic ele-
work, they are motivated to study nature’s
ments in architecture can counteract the increas-
shapes and compositional rules, and this can
ing dominance of functional semantic networks
enrich their creative curriculum.
and the associated epistemological attitude. Ad-
Besides having creative consequences, re-
mittedly, people will not consider biophilic ar-
duced contact with natural form could also sub-
chitecture or design as actual nature. However,
tly in?uence the way in which people think
such architecture shares some essential formal
about the world. Inquiries into semantic mem-
features with living things, and research indi-
ory indicate that processing conceptual infor-
cates that perceptual features are important for
mation about living things mainly relies on per-
recognizing living things. Biophilic design
ceptual information (e.g., the concept “zebra”
could lead to more attentiveness to an object’s
activates perceptual information, such as
perceptual qualities, thereby leading attention
“stripes”), whereas processing nonliving things
away from its possible functions and the asso-
or artifacts depends on functional information
ciated functionalist postures. Furthermore, be-
(e.g., the concept “knife” activates functional
cause of the (hardwired) emotional af?liation
information, such as “cutting”; e.g., Crutch &
with certain natural elements, nature-based ar-
Warrington, 2003; Farah & McClelland, 1991).
chitecture can awaken fascination for natural
These ?ndings could have important implica-
forms. Such an attitude could be ecologically
tions. The presence of nonnatural things, and
relevant, because it is found that proenvironmental
especially artifacts (e.g., cell phones, comput-
behavior is positively in?uenced by emotional
ers, chairs, pots, printers), is ever increasing in
states toward nature (Kals, Schumacher, &
the human living environment at the expense of
Montada, 1999).
natural structures or entities. A probable conse-
Without a doubt, people can get used to less
quence is that neural areas related to an object’s
formal diversity in the built environment. How-
functionality and hence functional analyses
ever, such a situation is not desirable because an
(i.e., how an object should be used or manipu-
increasing dominance of uniform (modernist)
lated) are becoming increasingly more domi-
environments will probably have a number of
nant in our thinking about the constituents of the
psychological and physiological costs. Recall
modern living environment. As we become
how, under evolutionary pressures, natural
more acquainted with such thinking, it is not
forms and environments became associated
implausible that it will be deployed in other
with a broad range of emotions, ranging from
domains as well (e.g., to generate explanations).
fear to excitement. In the human ancestral
This could especially occur when knowledge
world, such associations promoted ?tness be-
about phenomena in a certain domain, such as
cause they motivated the organism to undertake
the natural world, becomes increasingly more
adaptive reactions (e.g., ?ight). Today, there
scarce or more underdeveloped.
seems to be a discrepancy between the habitats
Although functional postures are important
humans have evolved in and modern urban set-
and necessary in certain ?elds, transferring
tings. For example, it was already noted that the

---

ARCHITECTURAL LESSONS FROM ENVIRONMENTAL PSYCHOLOGY
311
former was characterized by, among others, a
terventions are denoted as nature-based, or bio-
mix of complexity and order (S. Kaplan, 1987,
philic, architecture.
1988; Ulrich, 1983). Yet current architectural
settings do not appeal to this ordered complex-
Integrating Structural Landscape Features
ity. Modernist architecture mainly consists of
in Architecture
simple volumetric forms and thus deprives the
senses in their constant search for meaningful
How can the structural landscape features,
information. On the other hand, postmodern and
discussed earlier, be meaningfully applied to the
deconstructive architecture deliberately destroy
built environment? This is a more dif?cult issue
architectural coherence, either by jumbling
than applying well-de?ned natural contents to
together disparate stylistic and formal elements
architecture, because the former features are of
or by placing the destruction of coherence
a more abstract nature. Furthermore, only very
and structure at the heart of the tradition
few researchers have addressed this issue and
(Salingaros, 2004). Furthermore, modern build-
proposed clear guidelines on how to success-
ing is often dictated by ef?ciency and economic
fully integrate these qualities in architectural
motives, barely leaving room for symbolic and
settings.
stylistic references to natural contents (e.g., or-
First, turn to the type of setting that contains
nament; Pinker, 2002; Salingaros, 2004). In
an ideal mix of these structural landscape fea-
short, much of the modern built environment
tures, namely the savanna. An evident strategy
fundamentally lacks (references to) the contents
to imitate savannas is to integrate photographs
and structural organization that are characteris-
or projections of savannas in (interior) spaces.
tic of a good habitat. Exposure to such environ-
Another, more architectural method consists of
mimicking key structural features of savannas.
ments could rapidly and automatically trigger
Possible strategies include creating wide and
negatively toned feelings and the associated
open spaces; making variations in the architec-
stress-related endocrinal reactions (Ulrich,
tural topography; integrating clusters of real or
1983). Although such responses could go by
symbolic trees (e.g., columns); and integrating a
largely unnoticed because of human habituation
water feature (e.g., a fountain) or even a small
to this type of environment, the long-term oc-
?re. Note how certain retail settings, such as
currence of such stress reactions could have
shopping malls, often contain these elements.
important health effects (Parsons, 1991).
Because a major goal of the retail sector is
Increasing urbanization undoubtedly has a
attracting people, it should be no surprise that
number of positive consequences. For example,
organizational features of preferred settings are
in modern cities, people come to live closer
(intuitively) deployed in such commercial con-
together, which could promote social interac-
texts (Heerwagen, 2003).
tion and the pleasure and enjoyment associated
Because of their openness, savannas provided
with this (Van den Berg, Hartig, & Staats,
good prospects on the surrounding area. Fur-
2007). Furthermore, there is nothing inherently
thermore, trees typical of savannas (acacias)
wrong or undesirable about modern building
have low trunks and could, therefore, be
styles, and there is no reason to doubt the gen-
climbed to see across the landscape and to es-
uineness of positive reactions to such buildings.
cape predators. On the other hand, the broad
What, according to the current argument, could
canopies provided good protection against sun
become problematic is the (growing) domi-
and rain. Grant Hildebrand (1999) uses Apple-
nance of such nonnatural building styles at the
ton’s (1975) notions of prospect and refuge as
expense of settings with natural form languages
explanatory principles for the aesthetic appeal
(albeit natural or arti?cial ones). The core argu-
of certain buildings. Although Hildebrand does
ment of the current study is that, by including
not provide exact guidelines, his analyses show
elements of ancestral habitats in the built envi-
which spatial organizations in?uence the pros-
ronment, one can counter potential deleterious
pect and refuge dimensions of buildings. With
effects, which stem from this dominance, result-
regard to Frank Lloyd Wright’s house in Talie-
ing in more positive affects and more relaxed
sin, Wisconsin, Hildebrand notes the following:
physiological and psychological states. In the
“Deep overhanging eaves, alcoves and recesses,
remainder of this article, such architectural in-
the withdrawal of the house in the dense foliage,

---

312
JOYE
Figure 1.
Patterns can be obtained by some simple mathematical transformations (Salinga-
ros, 2003): (A) randomness; (B) translational symmetries; (C) re?ectional symmetries; (D)
rotational symmetries that are nested. (Copyright © Yannick Joye.)
and the cave-like masses of stone anchoring the
tion. Yet organic architects often use a geomet-
house to the hill all convey that this is a haven
ric module (e.g., a triangle) as main composi-
within which one can withdraw secure. Exten-
tional element (Mead, 1991). In this way, dif-
sive bands of window and the balcony reaching
ferent parts of the building are given a similar
out over the falling landscape, moreover, indi-
form, which results in an overall coherence
cate that the advantages of generous prospect
(Eaton, 1998).
are likely to be available within” (p. 28). It is
It could be noted that merely repeating sim-
clear that feelings of prospect and refuge can be
ilar elements does not guarantee an ordered
evoked by speci?c architectural interventions.
complexity. On the contrary: It can even lead to
Strategies for evoking concealment are reduc-
random structures, as in deconstructivism. A
ing lighting conditions, lowering ceilings, and
possible solution is to organize these (similar)
making small windowless spaces enclosed by
elements through patterns. These are often the
thick walls. The prospect dimension depends on
result of only a few simple mathematical oper-
opposite characteristics: larger space dimen-
ations, such as re?ectional, rotational, transla-
sions, raised ceilings, thin transparent walls,
tional, and glide symmetries. More complex
wide views on surrounding spaces, building on
patterns are obtained when these symmetries are
an elevated site, increased lighting conditions,
repeated or when they are nested (Salingaros,
balconies, and so on.
2003; Figure 1A–D). Traditionally, patterns
Prospect and refuge can be linked to the
take in a prominent place within the organic
predictors complexity and coherence, central to
tradition. Historically, they can also be found in,
the Kaplans’ preference matrix (e.g., R. Kaplan
for example, tiling, ornaments, mosaics, stained
& Kaplan, 1989). Only a setting that contains
glass windows, and (oriental) carpets.
enough prominent landscape features (e.g.,
Another structural feature that positively cor-
trees, rocks) can provide opportunities for ref-
relates with landscape selection is mystery.
uge. On the other hand, if a setting contains too
Some claim that this property can be conveyed
many elements, this makes it dif?cult to have a
by speci?c design elements: “When appearing
clear view over the landscape. Although com-
around corners, attached to walls, and hung
plexity and coherence have primarily been ap-
from ceilings, interesting objects, architectural
plied to landscapes, there is empirical evidence
details or motifs, graphics, video displays and
that a balanced presence of both properties con-
artifacts can create a little mystery and surprise
tributes to the aesthetic qualities of built settings
in the workplace” (Hase & Heerwagen, 2000, p.
(e.g., Herzog, Kaplan, & Kaplan, 1982). How-
30). However, the most straightforward way to
ever, how can a complex set of architectural
apply mystery to an architectural setting is by
elements be ordered? Again, take a look at the
de?ected vista. This can be realized by letting
architectural tradition associated with Frank
the architectural trail (e.g., corridor) bend away,
Lloyd Wright, namely organic architecture. In
which can lead to curiosity of what might lie
essence, organic architecture is not restricted by
beyond the bend, thereby encouraging explor-
stylistic conventions but is characterized by an
ative behavior. Another mode of mystery is
inherent form freedom. Although not necessary,
called “enticement.” Essentially, this notion re-
this often translates in buildings that are quite
fers to the situation in which a person is in the
irregular and complex both in plan and eleva-
dark, from where it can see a partially visible

---

ARCHITECTURAL LESSONS FROM ENVIRONMENTAL PSYCHOLOGY
313
and enlightened area or setting. Such enlight-
vegetal patterns, in traditional ornament. Ad-
ened regions draw attention and trigger explor-
mittedly, it could be possible that such imita-
ative behavior. Although mysterious settings
tions will not be very successful, because the
can be aesthetically appealing, too much irreg-
associated emotional states could quickly be
ularity or surprise can have the result that the
followed and suppressed by higher order or
layout of the building becomes confusing and
cultural beliefs. For example, the architectural
nontransparent, ultimately leading to orientation
community could consider such imitations as
and way-?nding problems. Legibility can be
kitsch. Nevertheless, it should be noted that
enhanced by integrating signalizations and dis-
there is often a discrepancy between what is
tinctive markings, by offering views on the out-
found appealing by experts and laypersons. The
side, and by making the building shape more
primary goal of this study is not to argue for
regular (Evans & McCoy, 1998).
what is supposedly fashionable or to defend
high art but to indicate what could be psycho-
Imitating Natural Contents in Architecture
logically appealing for the broad public.
An alternative to literal imitations is to create
How could natural contents be integrated into
architectural designs based on schematic imita-
the built environment? Evidently, this can be
tions of natural elements. These would no
done by providing views on the outside envi-
longer be exact copies but artistic interpreta-
ronment, by integrating vegetation in built set-
tions that still contain some global visual simi-
tings, by hanging nature pictures on the wall, by
larities with regard to the original natural object.
nature-oriented screensavers, and so on. These
One of the central claims of the current study is
interventions are what Stephen Kellert (2005)
that such constructions will be accompanied by
called indirect experiences of nature, and they
affective states that are similar to those evoked
come quite close to the design interventions
by real natural contents. Orians and Heerwagen
from the ?eld of evidence-based design (Ulrich
(1992, p. 572) expressed it as follows: “An
& Zimring, 2004; Van den Berg & Van Winsum-
evolutionary-ecological approach to aesthetics
Westra, 2006). It is in this sense that the modern
suggests that the incorporation of trees and tree
built environment sometimes imports some of the
forms, actual or symbolic, into the built envi-
icons of habitability that are typical of ancestral
ronment should have a strong positive impact
habitats. The result is that even architecture that is
on people. . .We predict that the presence of
characterized by nonnatural forms can be consis-
these ‘symbolic trees’ is associated with posi-
tent with the current argument. For example, some
tive response to built environments.”
modernist architecture (e.g., Mies’s Farnsworth
Although the occurrence of biophilic re-
House) is characterized by large expanses of glass,
sponses to symbolic representations of nature
by which the building opens up to the surrounding
could be prima facie plausible, it is problematic
natural landscape, potentially causing biophilic re-
that it is often taken for granted in the literature
sponses in the inhabitants.
on biophilic architecture. Although research
Implanting a building in a natural landscape
that has directly tested this prediction is lacking,
does not necessarily tell us something about the
some indirect arguments can be presented that
architectural form and whether it in some sense
support the conclusion. First, it is evident that
displays key features of our ancestral habitats.
domain-speci?c mechanisms in the brain will
Because we factually inhabit contexts in which
be activated by the objects in which they are
buildings are often more dominant than nature,
specialized. For example, a face detection
it also becomes relevant to come to biophilic
mechanism will be activated by its proper input:
interventions that pertain to the architectural
actual human faces. Yet it seems that such do-
form. But how should such interventions be
main-speci?c mechanisms do not care about
conceived? A ?rst strategy is to architecturally
whether the objects it analyzes are in any sense
imitate preferred natural entities, such as vege-
real or symbolic. More speci?cally, these neural
tative elements. Such imitations can take on
areas also tend to become activated by elements
different levels of abstraction. A ?rst option is
that share some central geometric features with
to literally copy these elements in architectural
the proper input of the domain-speci?c systems.
design. As already noted, there is an age-old
This is one reason why a symbolic representa-
tradition to copy nature, especially ?oral and
tion of a face, such as, for example, a smiley

---

314
JOYE
face ( ) or the front of a car, can be perceived
of ethology, such stimuli are labeled “supernor-
as having facelike features (Pinker, 1997; Sper-
mal stimuli” (Tinbergen & Perdeck, 1950).
ber & Hirschfeld, 2004) and can lead to the
Ramachandran and Hirstein (1999) argue how
onset of similar emotions as real eyes or faces
this phenomenon is one of the central laws that
(e.g., Aiken, 1998a). Similarly, it is probable
artists (unconsciously) deploy in art. They clarify
that the neural mechanisms specialized in pro-
this principle by referring to a phenomenon ob-
cessing natural elements will also be activated
served in the ?eld of animal discrimination,
by stimuli that share essential geometric fea-
namely the peak shift effect. A rat that is taught to
tures with natural elements, such as symbolic or
discriminate between a square and a rectangle and
imitative representations of nature in architec-
is rewarded for discriminating the rectangle will
ture. Because of the importance of quickly dis-
respond more frequently to the rectangle. How-
playing adaptive behavior to natural stimuli
ever, when the original rectangle is elongated, the
(e.g., exploration, escape, ?ghting), it is proba-
rat will respond even stronger to this new rectan-
ble that at the early stages of processing some
gle than to the rectangle that it was taught to
affective processing or priming will already
discriminate. According to Ramachandran and
take place, before any conscious recognition of
Hirstein (1999), artwork often taps a similar ef-
the imitated natural elements occurs (Ulrich,
fect: “What the artist tries to do (either con-
1983).
sciously or unconsciously) is to not only capture
Further reasons why architectural imitations
the essence of something but also to amplify it in
of nature could trigger biophilic responses are
order to more powerfully activate the same neural
more empirical in nature. First, it can be pointed
mechanisms that would be activated by the orig-
out that research on environmental preferences
inal object” (p. 17). According to Ramachandran
often uses simulations of nature (e.g., photos,
and Hirstein, such ampli?cations can occur along
posters, videos, and even paintings). The results
different dimensions of the artistic work: for ex-
that are obtained with these stimuli are close to
ample, form, color, and movement (see also
the responses associated with real nature, which
Aiken, 1998b). It is clear that, in the present dis-
suggests that realness does not play a decisive
cussion, main interest goes to ampli?cation of the
role. (Yet it should be noted that in such con-
architectural form, which can cause a peak shift
texts nature is mostly depicted very realistically,
effect with regard to real natural forms.
and when only realistic representations of na-
Maybe the most well-known examples of ar-
ture can be used in architecture, this restricts the
chitecture in which schematic interpretations of
range of possible architectural interventions al-
natural forms are present have been created by
most exclusively to ornamentation). Second,
Anton?´ Gaud?´. For example, the interior col-
symbolic representations of nature have been
umns of the Sagrada Familia are quite similar to
used throughout the history of art for aesthetic
treelike and ?owering structures (Figure 2A–B).
enhancements, which suggests that these can
Indeed, one can clearly differentiate a stem,
trigger biophilic responses. Third, research in-
which bifurcates into further branches and sub-
dicates that preferences for natural settings can
branches. The canopy of these treelike struc-
be statistically predicted by underlying geomet-
tures consists of ?owering forms, which further
ric characteristics, which lends plausibility to
strengthens the impression of symbolic vegeta-
the claim that geometric abstractions from na-
tion. A more modern architect whose work also
ture can cause the associated affective effects
contains schematic interpretations of natural ob-
(Ha¨gerha¨ll, Purcell, & Taylor, 2004; see also
jects is Santiago Calatrava. Like in Gaud?´’s
Fractals and Biophilic Reactions: A Critical
work, structural forces are an important deter-
Evaluation section).
minant of the shape of Calatrava’s architecture.
There is a further important reason why sym-
Yet he also seems to be directly inspired by the
bolic or schematic interpretations of naturalness
shapes of nature. According to Von Moos, “His
can lead to aesthetic reactions. This conclusion
architecture relates to the morphologies of plant
is based on the ?nding that formal abstractions
and animal life— on land, in the depth of the
or simpli?cations of certain conspicuous traits
sea, or in imagination” (Tischhauser & Von
of (survival-relevant) stimuli lead to similar,
Moos, 1998, p. 338). Particularly relevant for
or even stronger, emotional responses as the
the present discussion is that several building
original (natural) stimulus. Within the ?eld
elements resemble vegetative structures. For in-

---