Gifford Rrobet. Light, Decor, Arousal, Comfort, and Communication. Journal of EnvironmentalPsychology No. 8 (1988) P.P. 177-189.
Gifford Rrobet. Light, Decor, Arousal, Comfort, and Communication. Journal of EnvironmentalPsychology No. 8 (1988) P.P. 177-189.
Gifford Rrobet. Light, Decor, Arousal, Comfort, and Communication. Journal of EnvironmentalPsychology No. 8 (1988) P.P. 177-189.
LIGHT,
DECOR, AROUSAL,
COMFORT
AND COMMUNICATION
ROBERT GIFFORD
University o f Victoria, Victoria, British Columbia, Canada V8 W 2 Y2
Abstract
The effect of lighting level and room decor on interpersonal communication was
investigated. Arousal and comfort models were invoked to generate hypotheses that
(a) brighter lighting would stimulate more general communication, (b) lower lighting
levels would encourage more intimate communication, (c) over time, lower light
levels would dampen both general and intimate communication, and (d) home-like
decor would encour/tge more general and more intimate communication. In a 2 x 2
between-subjects design, pairs of female friends wrote two letters to one another in
bright vs. soft lighting and office-like vs. home-like decor. All the hypotheses were
confirmed, except that brighter light encouraged more rather than less intimate
communication.
Introduction
178
R. Gifford
psychologists have long since abandoned naive determinist views that fail to posit
plausible mediating factors between physical-setting variables and human behavior.
Thus, a second major goal of this study was to initiate research on light and decor
that advances a theoretical perspective, including suggestions as to mediators
between environmental variables (light and decor) and behavioral variables
(communication).
Almost no research except for the studies by Chaikin et al. (1976) and Sommer
(1969, 1974) has examined the effects of lighting on interpersonal communication.
For theoretical ideas, we were forced to start with animal research. The evidence on
light and activity at first appeared contradictory. Increased light seems to act as an
arousal agent in monkeys (Alexander and Isaac, 1965), but not in cats (Isaac and
Reed, 1961) or rats (Isaac and Kendall, 1967). However, this discrepancy has an
obvious explanation: monkeys are typically diurnal, rats and cats typically nocturnal.
A reasonable thesis is that for humans, as for monkeys, light usually serves as a
general arousal agent. Thus, human activity levels should generally increase in daytime or in bright light; as light dims, activity should slow. The invention of artificial
light has enabled humans to be quite active after nightfall, but we remain 'diurnal'
in that we still usually choose reduced lighting for relaxation and darkness for
sleeping. Recent work by Munson and Ferguson (in Gifford, 1987) suggests that, at
least for children, some forms of light appear to be more physiologically arousing
than others. Perhaps, then, bright light acts as an arousal agent, stimulating a variety
of behaviors, including amount of communication.
Carr and Dabbs (1974) found that dim lighting lessens eye contact and increases
verbal latency in conversations. Amount or quality of communication was not
studied. However, lowered eye contact and increased verbal latency suggest that
subjects spoke more intimately, which is consistent with the results of Chaikin et
al.'s findings for their 'soft' room. The 'soft' room, with its indirect lighting,
presumably was illuminated at lower levels than the 'hard' room, although Chaikin
et al. do not report illumination measurements.
Before specific hypotheses are advanced, the concept of interpersonal communication requires some differentiation. One person may communicate much information to another, but its content may be impersonal or only superficially personal.
This low-intimacy form of interaction may be termed 'general communication'.
Another person may communicate quantitatively little information, but its content
may be very personal and intimate. In the present study, the term 'intimate communication' is reserved for this high-intimacy form of interaction. An arousal model
would predict that general activity (in this case, general communication) increases
when a person is more aroused, but not necessarily that the content of the output
(intimacy of the communication) would increase. Lower light levels are associated,
in some contexts, with an increase in intimacy. More expensive restaurants and the
bedroom are two examples.
Based on the results of the studies described above and on the arousal model,
three hypotheses concerning illumination are advanced. First, brighter lighting produces more general communication. Second, low light produces more intimate
179
Decor has also been the subject of very few behavioral studies since the classic work
of Maslow and Mintz (1956), which examined interpersonal perceptions rather than
communication. Mehrabian and Diamond (1971) did investigate the efficacy of
conversation pieces. Although such objects apparently did enhance affiliative
behavior, examination of the regression equations in that study shows t h a t the
obtained effect was both complex and weak. One theme of this study and those of
Chaikin et al. (1976) and Sommer (1969, 1974) is that comfortable, pleasant settings
encourage more social interaction. Based on this comfort model, it seems reasonable
to tentatively hypothesize that a 'soft' (more home-like) setting will evoke both
more intimate communication and more general communication than will a 'hard'
(more office-like) setting.
Method
Subjects
Female undergraduates in an introductory social science class were recruited by
telephone on a voluntary basis (no payment in money or course credit) for 'a study
of conversations'. It was decided to s t u d y communication in mature relationships
rather than between strangers meeting for the first time, and women who volunteered
were asked to bring a friend to the study. Thus, the participants were those who,
since they already knew and liked one another, might be expected to exchange
information with one another, in their daily lives. This procedure resulted in the
participation of 72 women in the 18 to 25 age range.
When pairs arrived, they were assigned to one of four experimental conditions
according to a rotating, predetermined order (i.e., condition 1, 2, 3, 4, l, 2, etc.).
The four conditions were those resulting from the combination of two levels of
lighting and two levels of decor. These assignments were essentially random because
the subjects suggested their own appointment times during the telephone call; neither
they nor the experimenter could have known which appointment time would, once
the schedule was complete, result in assignment to which condition.
The settings
An attempt was made to create lighting and decor conditions that might actually be
found in natural settings. The bright light condition consisted o f the normal
fluorescent lights installed in the laboratory room when it was built a few months
before the study began. Six ceiling-mounted 35-watt bulbs illuminated the 2.25 m
3.55 m room. In the soft condition, one 100-watt bulb in a table lamp was lit. In the
latter case, the lighting, although soft, was quite sufficient for writing without
eyestrain. In physical terms, the bright condition averaged 90 footcandles at the two
positions and the soft condition averaged 6 footcandles at the two positions, based
on measures from a Canadian Research Institute photometer (model D P M T - I C C )
that were color- and cosine-corrected for human perception.
In the office decor condition, typical modern office furniture, unembellished by
180
R. Gifford
any personal touches, was used. Participants sat on armless molded-plastic chairs
and the experimenter sat at a standard desk. The experimenter's desk was nearly
covered with neat piles of paperwork. Textbooks were displayed on top of a plain
metal file cabinet. A typewriter and the table lamp sat on a small table. The floor
was covered with the office carpeting supplied when the r o o m was built. A university
calendar was the only wal.1 adornment. In sum, the office was furnished in a standard
way, tidy but somewhat bare and impersonal.
For the home-like decor condition, we attempted to create an office that might be
found in the possession of a person who had made an effort to soften the institutional
quality of the space. The same r o o m was used and the desk and file cabinet were
retained. The changes included adding a potted plant, hanging two art posters on
the wall, placing a small decorative rug on the standard carpet, removing the
typewriter, placing a tablecloth under the table lamp on the small table, and adding
an antique wooden radio to the experimenter's desk, which was now clear of papers.
The participants sat in padded armchairs instead of molded plastic chairs.
In all conditions the participants sat about 30 degrees from a face to face position
at a distance of 1-5 m (chair center to chair center). Placed before each participant
was a music stand that was adjusted to an angle and height suitable for writing.
Procedures
Upon arrival, the participants took their places and the experimenter explained that
we were interested in the interactions between two people in an ongoing relationship;
that we wished to learn about how people choose to let friends and acquaintances
know about themselves.
A written-communication procedure (cf. Worthy et al., 1969; Rubin, 1975; Archer
and Berg, 1978) was chosen in favor of oral communication. We wished to avoid
forcing the participants to tailor their statements to suit the presence of a thirdparty stranger, the experimenter. The subjects were all friends, but none of them was
acquainted with the experimenter. Thus, a written communication approach combined with assurances of confidentiality presumably allowed fuller, freer and more
intimate communication between friends than would oral communication with the
experimenter present, or written communication with the expectation that the
experimenter would be reading the material later.
Participants were told they would write two letters to their partners, but the
letters would be delivered immediately after the session rather than being mailed.
The participants understood and readily accepted that the purpose of asking them
to write to one another instead of holding a conversation was to ensure the privacy
of the communication between them. The participants were assured that the experimenter would never know the contents of their disclosures to one another.
Next, the participants were shown a list of 11 suggested topics for their letters (see
Table 1). Most of the topics were selected from a longer list first used in disclosure
research by Rubin and Shenker (1978). Some were adapted slightly and a few new
ones were added, to ensure a broad range and depth of topics.
Participants were asked to select their first topic by blindly picking from an
envelope slips of paper that had topic numbers written on them. To enable a
comparison of participant reponses on a c o m m o n topic, all the slips requested the
participant to write on topic 11, 'The things that I worry a b o u t the most'. They then
wrote a letter on this topic to their partners. To allow a choice o f topics in the second
181
letter, all the slips said, 'Choose any topic except the one you have already written
about.'
No time limit was set for writing on either topic. Participants placed their completed letters in envelopes and were given the opportunity to exchange letters with
their partners at the end of the session. The experimenter never saw the contents of
the letters.
Measures
Variables in the study included, in order of administration: (a) questionnaire items
about the relationship between the partners; (b) participant ratings of the intimacy
of the 1 l topics; (c) two measures of general communication and two measures of
intimate communication; and (d) questionnaire items used to check the manipulation
of the independent variables. Each of these will, in turn, be described more fully.
During the introduction phase of the study participants responded to a brief
questionnaire containing three questions: 'How long has it been since you first knew
or heard of your partner?' 'If there was any significant gap between the time you
first knew of your partner and the time you first significantly interacted (i.e., spoke
more than a brief greeting to each other), how long was this gap?' 'How would you
characterize your present relationship?' For the latter item, seven labelled alternatives from 'Someone I try to avoid' to 'My closest friend' were offered.
Next, participants rated the intimacy of all eleven topics on 1 to 8 bipolar scales
('not intimate' to 'extremely intimate') at the time they were familiarized with them.
These ratings were obtained to allow later analyses of the intimacy of topics selected
by the subjects.
General communication was assessed by two measures, the total number of words
written in each letter (total words) and the length of time each participant spent
writing each letter (writing time). One measure of intimate communication, designed
to respect the privacy of the interaction between friends yet encourage full and
frank communication, was the total number of selJ:referent words used (e.g., I, me,
mine, myself).
To maintain the confidentiality of the disclosures, the two word counts (total
words for general communication and self-referent words for intimate communication) were made by the participants themselves after they had written both letters.
T h e other general communication measure, writing time, was covertly measured by
the experimenter.
The second measure of communication intimacy combined the intimacy of the
topic and the intimacy of one's revelations on that topic. This measure, intimacy,
was computed as the product of the topic's intimacy as rated by the participant and
her response to the following questionnaire item: 'How intimate or revealing were
your statements on this topic?' On both items, response categories varied from 1,
'Not at all' to 8, 'Extremely'. Thus, the intimacy score would be 30 for a topic
judged by a participant to have an intimacy level of 5 with the participant's own
disclosure intimacy on that topic judged by her to be 6.* The participants also
reported which topics they had written on. This response was, of course, relevant
* The results of a M A N O V A using only the participants' ratings of the intimacy of their disclosures,
without the topic intimacy component, were very similar to those obtained in the M A N O V A reported in
the results section.
182
R. Gifford
only for the second letter, but it was requested for both letters because participants
believed that their first topic selection was random.
Finally, to check the manipulation of the independent variables, participants
reported their assessments of the room at the end of the session. The two assessments
of interest, of lighting and decor, were embedded among two others, of room size
and aesthetics. The response categories for the decor items ranged from 1, labelled
'very informal, homey decor' to 8, 'very formal, office-like decor'. Response categories for the lighting item ranged from 1, 'Very soft lighting' to 8, 'Very bright
lighting'.
Results
183
TABLE 1
Mean ratings of topic intimacy
1
2
3
4
5
6
7
8
9
10
11
3.12
5.39
3.75
6.77
2.99
5.32
6.53
3"07
5.25
4.36
5.82
Note: Ratings are based on an 8-point bipolar scale on which 1 = not intimate and
8 = extremely intimate.
TABLE 2
Communication means by condition and letter
Bright light
Soft light
Home-like
Office-like
Home-like
Office-like
Letter 1
Total words
Self-referent words
Writing time (min)
Intimacy
249.50
23.50
13.14
44.17
244.60
28.22
11.59
32.89
241.70
27.27
13.95
32.50
212.20
24.72
11.75
34.11
Letter 2
Total words
Self-referent words
Writing time (min)
Intimacy
252.20
21.72
13.61
44.33
252.20
24.56
12.35
28.44
209.50
18.89
10.26
24.00
209.30
21.67
9.45
23.83
Note: Intimacy scores are the product of two intimacy ratings, one for the topic itself and one for the
participant's communication on the topic. Both are 1 8 scales with 8 = extremely intimate. Thus, higher
scores indicate greater intimacy.
184
R. Gifford
spent 12.7 minutes writing in the home-like decor and 11.3 minutes writing in the
office decor (P = 0.06). Intimacy scores were greater in the home-like decor (36.2)
than in the office decor (29.9; P = 0.08). Thus, the hypothesis that more intimate
communication and more general communication occurs in home-like settings is
modestly confirmed for two dependent variables.
The significant effect for light [multivariate F(4, 65) = 2.36, P = 0.06] is manifested on the same two univariate measures. In both cases, more communication
occurred in brighter light (12.7 vs. 11.4 minutes, P = 0.09, Intimacy scores of 37.5
vs. 28.6, P < 0.02). These results confirm the hypothesis concerning general communication, but are significantly opposite to the predictions made about intimate
communication.
In sum, the results reported so far indicate that more intimate and more general
comrfiunication occurs in home-like and bright conditions. Table 3 illustrates these
trends.
TABLE 3
Mean writing time and intimacy by light and decor
Decor
Light
SoJr
Writing time (rain)
Intimacy
Bright
Writing time (rain)
Intimacy
Home-like
Office-like
12"1
28"2
10-6
29-6
13'4
44"2
12"0
30'6
Discussion
The results clearly show that lighting and decor affect general and intimate communication. They are generally consistent with the proposals that arousal mediates
o Bright light
Soft light
14
WRITING
TIME
(min.)
185
13
12
11
10
2
LETTER
o Bright light
Soft light
40
35
INTIMACY
30
25
20
LETTER
FIGURE1. The decline of writing time and intimacy from letter one to letter two in soft lighting, compared
to bright lighting.
the effects of light and that comfort mediates the effects of decor. The first hypothesis, that more general communication would occur in brighter settings, was
confirmed. Second, more intimate communication was predicted for lower light
levels. This was significantly disconfirmed; more intimate communication occurred
in the brighter light than in the softer light. Third, communication declined in softer
light from letter one to letter two, as predicted. Fourth, general and intimate communication were hypothesized to be greater in home-like conditions; these hypotheses were confirmed.
In most of these cases, support for the hypotheses came from one of the two
measures of general and intimate communication. In the case of general communi-
186
R. Gifford
cation, writing time was always the significant variable; total words never was. In the
case of intimate communication, intimacy was always the significant variable; selfreferent words was significant in only one case. If nothing else, this indicates the
value of differentiating the concept of interpersonal communication and of the value
of using multiple measures. Why light and decor are related to writing time but not
total words is not clear. Intimacy, based on the participant's cognitive assessment of
the topic and her own letter, is frequently related to light and decor while self-referent
words, a mere word count, is not. Perhaps the word count measures, neither of
which were very useful, are too subject to extraneous influences such as writing
style. At least in the case of intimacy, a more psychological measure seems to
be more sensitive to differences in light and decor than are the more mechanical
measures.
The similarity of effects on intimacy and writing time raises the issue of their
relatedness. They are conceptually different, but because they are similarly affected
by the independent variables, are they highly correlated? No; they correlate 0.31,
which is statistically significant, but they share less than 10% of their variance.
Lighting
One interesting outcome is the disconfirmation of the hypothesis predicting that soft
lighting would be associated with more intimate communication. Because this
hypothesis was derived from the Chaikin et al. study, a comparison of the studies
may help elucidate the reasons for the discrepancy. One difference is that in this
study participants judged the intimacy of their own disclosures, but in Chaikin et
al.'s study another person (an interviewer) did so. Second, this study examined written
communication but Chaikin et al. examined oral communication. Third, this study
focused on established friendships but Chaikin et al. focused on interactions of
strangers. Fourth, this study's participants were peers but in Chaikin et al.'s study a
subtle but important status or role difference was present in that the experimenter,
who was in her thirties, interviewed introductory psychology students who were
presumably about a decade younger; also, the experimenter played a counselor role
while the subjects played a counselee role. Any of these differences might have
accounted for the opposite findings. One task for future researchers is to discover
the conditions under which brighter light elicits more or less intimate communication.
The results for light are in accord with an arousal model. In the conditions of this
study, more light produced not only more general communication but also more
intimate communication.
The light letter interaction
As predicted, both intimate and general communication declined over time in softer
light. The stronger effect was for writing time. The results are consistent with an
arousal model that placing individuals in soft lighting will lower their arousal levels
which, in turn, will lower their activity levels.
Most university lighting is at levels closer to our bright condition than to our soft
condition. Thus, the arousal and writing time of participants in the soft lighting
condition appears to have decreased with lengthening exposure to soft light, rather
than that the arousal and writing time of bright-light participants increased with
lengthening exposure to bright light.
187
Decor
More home-like decor was associated with more communication of both kinds.
Homey decor probably is not only more physically comfortable, but it may also be
more psychologically comfortable, evoking a sense of refuge and unguardedness
that is associated with home. Both these feelings would be expected to encourage
individuals to communicate more, both generally and intimately.
The letter main effect
One unexpected finding was that communication declined from letter one to letter
two, not only in soft light as predicted, but regardless of light and decor. This may
have happened for three possible reasons. First, participants may simply have tired
of writing. However, total words did not change significantly from the first letter to
the second, and the significance of the decline in writing time, the other general
communication measure, was marginal. Second, the intimacy of what participants
wrote on the second topic may have been lower. A t-test did not support this notion,
either (P > 0-10). Third, the intimacy of the topic itself might have been lower for
the second letter. This was true: the intimacy of topic one was 5.82 compared to
4.76 for topic two [t (71) = 3.75, P < 0.001]. This difference between letter one and
letter two carried over into the intimacy scores, which were computed as the product
of these two ratings. Letter one's mean intimacy score was 35.9; letter two's was
30-6 [t (71) = 2-70, P < 0.01].
Thus, it appears that the decline in communication over time is primarily a decline
in the intimacy of what was written. This may have resulted from the choice of
topics. The first (prescribed) topic, 'The things I worry about the most', was the
third-most intimate of the 11 topics. At the outset of the study, topic intimacies
were not yet known, so the selection of an opening topic could not have been
guided by knowledge of the intimacy of this topic. The odds were, then, that a
second topic would be closer to the mean intimacy of the remaining 10 topics. In
fact, this mean was 4.66 and the mean intimacy of the remaining topics was 4.76.
Therefore, the intimacy decline over time was probably due to this topic selection
factor.
Design applications
In combination, the findings of this study suggest that if one wishes to increase both
general and intimate communication, the setting should be normal-bright (i.e., about
90 footcandles) and home-like in decor. Naturally, these results so far may only be
applied to female friends writing letters; broader generalization depends on research
with other forms of communication, males, and stages of relationship. However,
these results are sufficient to create caution about accepting the common notion
that more intimate communication will occur in dimly lit settings. At the same time,
they reinforce the notion that home-like settings encourage communication.
Conclusion
This study began as an attempt to clarify past studies which showed that changes in
communication were related to the physical setting, but were unable to specify which
188
R. Gifford
features of the setting were responsible. We posited that arousal and comfort mediate
written communication and the results are generally consistent with these proposals.
The findings suggest some obvious future directions for researchers. Some of
these were made salient in comparing this study's procedure with that of the Chaikin
et al. study: oral versus written communication, who judges communication
intimacy, interactions between friends versus strangers, and role or status differences.
Research also might be directed toward assessing bodily states during communication under different lighting conditions to determine whether physiological arousal
changes occur as the results of this study suggest. Whether or not the results are best
explained by arousal and comfort models, the study (a) confirms the hypotheses
that light and decor affect communication, and (b) places the research within a
theoretical context. Another question for the future is whether brighter-than-normal
lighting actually increases communication, in contrast to the present finding that
normally bright lighting maintains communication levels while subnormal lighting
levels diminish it.
Acknowledgements
References
Alexander, M. and Isaac, W. (1965). Effect of illumination and d-amphetamine on the activity
of the Rhesus macaque. Psychological Reports, 16, 311-313.
Archer, R. L. and Berg, J. H. (1978). Disclosure reciprocity and its limits: a reactance analysis.
Journal of Experimental Social Psychology, 14, 527-540.
Carr, S. J. and Dabbs, J. M. (1974). The effects of lighting, distance, and intimacy of topic on
verbal and visual behavior. Sociometry, 37, 592-600.
Chaikin, A. L., Darlega, V. J. and Miller, S. J. (1976). Effects of room environment on selfdisclosure in a counseling analog. Journal of Counseling PsTchology, 23, 479481.
Cozby, P. M. (1973). Self-disclosure: a literature review. Psychological Bulletin, 79, 73-91.
Gifford, R. (1987). Environmental Psychology: Principles and Practice. Boston: Allyn and
Bacon.
Isaac, W. and Kendall, N. P. (1967). Sensory stimulation and timing behavior. Psychonomic
Science, 8, 41~42.
Isaac, W. and Reed, W. G. (1961). The effect of sensory stimulation on the activity of cats.
Journal off Comparative and Physiological Psychology, 51, 172-174.
Maslow, A. H. and Mintz, N. L. (1956). Effects of esthetic surroundings: 1. Initial effects of
three esthetic conditions upon perceiving 'energy' and 'well-being' in faces. Journal of
Psychology, 41, 247-254.
Mehrabian, A. and Diamond, S. G. (1971). Effects of furniture arrangement, props, and
personality on social interaction. Journal of Personality & Social Psychology, 20, 18-30.
Rubin, Z. (1975). Disclosing oneself to a stranger: reciprocity and its limits. Journal of
Experimental Social Psychology, 11,233-260.
Rubin, Z. and Shenker, S. (1978). Friendship, proximity, and self-disclosure. Journal of Personality, 46, 1 22.
Sommer, R. (1969). Personal Space." The Behavioral Basis qf Design. Englewood Cliffs, NJ:
Prentice-Hall.
Sommer, R. (1974). Tight Spaces. Hard Architecture and How to Humanize It. Englewood
Cliffs, NJ: Prentice-Hall.
189
Won-Doornink, M. J. (1979). On getting to know you: the association between the stage of a
relationship and reciprocity of self-disclosure. Journal of Experimental Social Psychology,
15, 229 241.
Worthy, M., Gary, A. L. and Kahn, G. M. (1969). Self-disclosure as an exchange process.
Journal of Personality and Social Psychology, 13, 59-63.