Blue space: The importance of water for preference, affect, and restorativeness ratings of natural and built scenes

https://doi.org/10.1016/j.jenvp.2010.04.004Get rights and content

Abstract

Although theorists have suggested that aquatic environments or “blue space” might have particular restorative potential, to date there is little systematic empirical research on this issue. Indeed the presence of water has, unintentionally, been a confounding factor in research comparing people’s reactions to built and natural environments. Whereas aquatic features (rivers, lakes, coasts) are frequently present in visual stimuli representing natural environments they are rarely incorporated in stimuli portraying built environments. As many towns are, for good reason, located near water this is a potentially significant oversight. The current research collated a set of 120 photographs of natural and built scenes, half of which contained “aquatic” elements. Proportions of “aquatic”/“green”/“built” environments in each scene (e.g. 1/3rd, 2/3rds) were also standardised. Two studies investigated preferences (attractiveness, willingness to visit and willingness to pay for a hotel room with the view), affect and perceived restorativeness ratings for these photographs. As predicted, both natural and built scenes containing water were associated with higher preferences, greater positive affect and higher perceived restorativeness than those without water. Effect sizes were consistently large. Intriguingly, images of “built” environments containing water were generally rated just as positively as natural “green” space. We propose a number of avenues for further research including exploration of the mechanisms underlying these effects.

Introduction

The role fresh water plays in our physiological health is clear. We can survive only a few days without it. The role water plays in our psychological health is far less obvious. We know that aquatic environments were revered in many ancient societies (e.g. Egypt, Greece, Rome; Solomon, 2010) and that such reverence continues today (Ruback, Pandey, & Kohli, 2008). Many religions promote ritual washing and/or immersion and specific aquatic environments continue to have spiritual importance: For example, the Ganges in Hinduism, the Well of Zamzam in Islam and Lourdes in Roman Catholicism. At a secular level people are prepared to pay more for houses and hotel rooms with views of water (Lange and Schaeffer, 2001, Luttik, 2000), and aquatic environments are a frequent aspect of people’s favourite places (Kaplan and Kaplan, 1989, Korpela et al., in press), preferred leisure destinations (Natural England, 2009) and recollections of positive childhood activities (Waite, 2009). Yet despite the enormous cultural and economic importance of aquatic environments we still know very little about people’s perceptions of these environments in psychological terms.

From an evolutionary perspective the reverence of water makes sense. Early humans attracted to aquatic environments with supplies of fresh water may have been more likely to survive than those attracted to non-aquatic environments. Additionally, salt-water environments provided food and migratory corridors and it has been suggested that Omega 3 fatty acids present in large amounts in fish and shellfish may have played a crucial role in the development of the brain and the evolution of modern humans (Hardy, 1960, Morgan, 1997). Thus it seems reasonable to argue that we have evolved preferences for such environments (Kaplan and Kaplan, 1989, Ulrich, 1983). Perhaps this apparently obvious connection is one of the reasons why aquatic environments have been relatively under-researched in environmental psychology. Instead, studies investigating the actual or perceived restorativeness of different environments have typically focused on the natural vs. built distinction and only rarely examined the role of water within these two categories.

The aim of the current research was to help redress this imbalance. Employing the simple but widespread method of asking people to rate photographic scenes we systematically explored reactions to natural and built environments with or without aquatic elements. Following on from previous research we measured people’s preferences for the different scenes as well as their affective reactions and perceived restorativeness. Our main hypothesis was that people would respond favourably to both natural and built environments containing water. Moreover, since we controlled the amount of water visible in the scenes we were also able to examine possible dose-response relationships: Are people sensitive to the proportion of water in a scene?

For the purposes of the present research we consider an aquatic environment to be any environment, natural or urban, containing visible amounts of standing or running water (but not falling rain or settled ice or snow). Although aquatic environments might usually be considered to be ones dominated by water (e.g. seas and lakes) we also include environments where water may be a secondary aspect of the scene (e.g. fountains or streams; cf. Herzog, 1985). This enables the potential dose-response relationships between people’s preferences and ratings and the amount of water in a scene to be investigated.

Perhaps the most compelling evidence that people value aquatic environments comes from research into behavioural choices. Luttik (2000), for instance, reviewed house prices across several districts of the Netherlands and concluded that, all else being equal, people were prepared to pay between 8 and 12% more for houses with views of water. Lange and Schaeffer (2001) report similar findings for rooms in two hotels in Zurich. In one, rates for rooms overlooking the lake were roughly 10% higher than those with views to a forest. These data are supported by expenditure on leisure excursions to aquatic environments. In the UK, for instance, there are nearly 250 million visits to the coast and a 180 million to other aquatic environments such as rivers, canals and lakes annually (Natural England, 2009). However, these findings are indicative only because it may be that other features of these environments, e.g. their remoteness or lack of human construction rather than the presence of water per se, are driving these effects.

More direct, but also more equivocal, evidence comes from psychological research. First, Herzog (1985) examined preferences for different waterscapes and found that photographs of “swampy areas” were rated significantly less positively than “rivers, ponds and lakes”, “mountain waterscapes” and “large bodies of water”. Although suggesting variability in response to aquatic environments this study did not, however, examine non-aquatic environments so people’s relative preferences between the two could not be assessed. Nevertheless, the study did suggest a possible dose-response element to aquatic environments. Specifically, the highest preference ratings were for “large bodies of water” suggesting preferences might increase as the proportion of water increases.

A second study, by Ulrich et al. (1991), did compare people’s reactions to videos of natural scenes either with or without water present (in addition to urban environments). Instead of preference ratings, however, the researchers were interested in the degree to which being exposed to different environments following a stressful situation would help “restore” people’s physiological functioning, e.g. blood pressure and skin conductance, to pre-stress levels. Twenty participants were shown a video of “Nature Vegetation” and 20 of “Water”. The water video was described as a: “Setting dominated by trees and a fast moving stream; waves and ripples on stream surface. No people or animals. Constant 63–64 decibel sound from stream”. The authors state that they found no significant differences between the Natural Vegetation and Water settings (although no data are presented).

While this suggests that water may not have any particular restorative properties, a number of alternative interpretations are possible. First, the scene was “dominated by trees” making it possibly a green-space scene with some water rather than a “large body of water” with some green (cf. Herzog, 1985). Second, the average level of noise in the water video was higher (though the maximum was the same) and it may be that any positive effects from viewing water were counteracted by negative effects of a more noisy auditory stimulus. Third, there may have been insufficient power to detect differences with only 20 participants per condition. Fourth, only one scene was used making it hard to know whether this was representative of all aquatic environments and whether the effects are generalisable.

A third study, by van den Berg, Koole, and van der Wulp (2003) also directly compared the presence of aquatic elements in natural and urban settings. Participants were presented with videos of walks in Urban and Rural environments that either did or did not contain water. Similar to Ulrich et al. (1991) they found no effect of water on “affective restoration, concentration, beauty or naturalness” (p.141). A closer inspection of the videos reveals a potential explanation: The water in these scenes is a thin, dark brown, unattractive stream or canal. As the authors themselves recognise: “our manipulation may have been too weak to have any impact on participant’s preferences or restoration…[thus]… it seems worthwhile to further investigate the independent contribution of water to the occurrence of restorative effects in future studies” (p.145).

This continuing ambiguity concerning the impact of water on well-being has led some researchers to eliminate it from all scenes (e.g. Herzog, Maguire, & Nebel, 2003). Although this aids interpretation it does not help us understand the role that water plays. More commonly studies have included aquatic stimuli but not systematically (e.g. Berman et al., 2008, Berto, 2005, Pretty et al., 2005). In reviewing these papers our goal is not to be unjustly critical. The original stimuli were constructed for other purposes and within their frame of reference we find the results compelling. Nevertheless, we did find a systematic, though no doubt unintentional, bias towards the inclusion of aquatic scenes in the more positive categories of stimuli. For instance, while Berman et al.’s (2008) set of 50 nature photographs contained 39 with water (78%), not one of their urban ones did. Given that many towns and cities are built, for good historical reasons, near the coast, along rivers or on lakes (Solomon, 2010) and often contain human-made aquatic elements such as ponds or fountains (Nordh, Hartig, Hagerhall, & Fry, 2009) this appears to be a potential oversight.

A similar picture emerges with Berto’s (2005) data. Of her 25 Restorative images, 19 (76%) contained water while water was present in only 2 of 25 (8%) Unrestorative scenes. More informative still are Pretty et al.’s (2005) stimuli. These authors developed four categories of image along two dimensions: Rural/Urban crossed with Pleasant/Unpleasant. Participants in each of the four groups were presented with slide shows of 80 images. The number containing water was: Rural Pleasant = 31; Rural Unpleasant = 6; Urban Pleasant = 38; Urban Unpleasant = 6. Thus while 39% of Rural Pleasant and 47% of Urban Pleasant scenes contained water, only 8% of Rural and Urban Unpleasant ones did. We interpret this as support for the contention that water may play a key role in preferences and potential restorativeness for both Rural and Urban scenes.

Far from being isolated examples these findings reflect an underlying pattern in the literature. Of the 6 landscape posters used by Kweon, Ulrich, Walker, and Tassinary (2008) to reduce anger and stress at work, 5 were of aquatic scenes. In Nanda, Eisen, and Baladandayuthapani’s (2008) survey of art preferences among patients and students, aquatic scenes consistently ranked highest. Purcell, Peron, and Berto’s (2001) participants ranked “lake” scenes as the most preferable and restorative, if not most familiar, landscape. The highest ranked Forest scenes in Han’s (2007) research contained water. The examples in Weinstein, Przybylski, and Ryan (2009) paper are of a “natural” aquatic scene vs. an “urban” one with no water. Felsten (2009) noted that “natural murals, especially those with water, were more restorative” (p.160). Nordh et al. (2009) found that small urban “pocket parks” containing water were rated as more restorative, in terms of “being away” and “fascinating” than those without water. Finally, although not predicted, Karmanov and Hamel (2008) found an urban scene with a large amount of water was generally rated as positively as rural green space.

Moreover, previous data sets were not only less likely to include water in the less positive categories but when it did appear there tended to be less of it. For instance, the largest proportion of water in Berto’s (2005) Unrestorative scenes represented around 5% of the image while her Restorative scenes contained up to 50%. Similar proportions are evident in Pretty et al.’s (2005) Pleasant vs. Unpleasant stimuli, and some of Berman et al.’s (2008) Nature scenes contain around 60% water. We interpret these findings to suggest that there may be a dose-response effect with restorativeness increasing in proportion to the amount of water present in the scenes. The basic idea builds on Tennessen and Cimprich’s (1995) study which categorised views from student dorms as either, All Nature, Mostly Nature, Mostly Built or All Built.

Clearly, the idea that water may have restorative potential is not original (e.g. see Kaplan & Kaplan, 1995). Nonetheless, it seems that many studies have unintentionally supported, and in many cases not even recognised, this possibility. The central aim of the current research was to therefore explore this issue more systematically. We had three main objectives. First, and most importantly, we wanted to systematically examine whether the presence of water in both natural and built environments would have a positive impact on a range of preference and rating indicators. Second, by controlling the proportion of water in a scene we wanted to explore potential dose-response relationships: e.g. do reactions to scenes become more positive as the proportion of water increases? Third, we wanted to revisit the Green/Built relationship but submit this to more rigorous testing by excluding water from both environments. These objectives were addressed in two studies.

Measures of preference in Study 1 were attractiveness (e.g. Herzog, 1985) and willingness to visit the scene (an approach-avoidance measure, Ulrich, 1983), and in Study 2 willingness to pay (WTP) for a hypothetical hotel room with the specific view (Lange & Schaeffer, 2001). In terms of other ratings, in Study 1 we measured affective valence and arousal (De Kort, Meijnders, Sponselee, & Ijsselsteijn, 2006) and in Study 2 perceived restorativeness (Study 2; Berto, 2005, Hartig et al., 1997).

Earlier research demonstrated that Built environments containing Green space are rated more positively than those without. Extending this logic we hypothesised that as the proportion of visible Green space increases, so the rated positivity of the scene will also increase. This will manifest itself in stronger preferences for (attractiveness/willingness to visit/pay) and more positive ratings of (valence/arousal/perceived restorativeness) environments with more vs. less green space.

Our review above suggests that natural scenes containing water are more likely to be viewed positively. Extending this logic, we hypothesise that as the proportion of Aquatic space visible in natural scenes increases, so preferences and ratings will also increase.

If the first two hypotheses are correct then adding water to Built environments should also increase positivity towards them. Moreover, as the proportion of Aquatic space visible in Built scenes increases, so preferences and ratings should also increase.

Finally, if Hypotheses 1–3 are supported then including water in Built environments should result in greater positivity ratings than adding Green space. Specifically, we hypothesised that preferences and ratings of Built scenes containing Aquatic elements will be more positive than those containing Green elements.

Section snippets

Preparatory work: development of photo sets

We began by identifying three archetypal environments: “Aquatic” including water and associated aspects (e.g. rocky shore, sandy beach); “Green” including fields, forests, parkland; and “Built” including buildings and constructions (e.g. walls, roads). We then collated images that either reflected a single environment (e.g. “Green-Only”) or a combination of two (e.g. “Green-Aquatic”). The ratio of these elements in mixed environments was crucial. Following the “rule-of-thirds” in photography (

General discussion

Our review of the literature comparing people’s reactions to different environments suggested an unintentional bias with water featuring more often, and in larger quantities, in natural than built environments. Since water may be a positive environmental feature, which for historical reasons is a central aspect of many towns and cities, this may have been a potentially important oversight. We therefore collated a large set of scenes (N = 120) that systematically controlled for the presence of

Conclusion

Despite the fact that water covers more than two-thirds of our planet, systematic investigation of relationships with aquatic environments, or Blue Space, has been neglected. The current findings suggest this is an important oversight because these environments were associated with higher preferences and more positive subjective reactions than both natural and built settings without water. Although replication of our findings using different techniques and exploration of the underlying

Acknowledgements

We would like to thank Lynne James for help with programming, Rita Berto, Jo Barton and Marc Berman for further information on their inspiring studies and three anonymous reviewers for their excellent comments on a previous version of this manuscript.

References (56)

  • T.R. Herzog et al.

    Assessing the restorative components of environments

    Journal of Environmental Psychology

    (2003)
  • S. Kaplan

    The restorative benefits of nature: toward an integrative framework

    Journal of Environmental Psychology

    (1995)
  • D. Karmanov et al.

    Assessing the restorative potential of contemporary urban environment(s): beyond the nature versus urban dichotomy

    Landscape and Urban Planning

    (2008)
  • E. Lange et al.

    A comment on the market value of a room with a view

    Landscape and Urban Planning

    (2001)
  • G. Loomes

    (How) can we value health, safety and the environment?

    Journal of Economic Psychology

    (2006)
  • J. Luttik

    The value of trees, water and open space as reflected by house prices in the Netherlands

    Landscape and Urban Planning

    (2000)
  • R. Miceli et al.

    Disaster preparedness and perception of flood risk: a study in an alpine valley in Italy

    Journal of Environmental Psychology

    (2008)
  • H. Nordh et al.

    Components of small urban parks that predict the possibility of restoration

    Urban Forestry and Urban Landscaping

    (2009)
  • R.B. Ruback et al.

    Evaluations of a scared place: role and religious belief at the Magh Mela

    Journal of Environmental Psychology

    (2008)
  • H. Staats et al.

    Alone or with a friend: a social context for psychological restoration and environmental preferences

    Journal of Environmental Psychology

    (2004)
  • P. Suedfeld et al.

    Water immersion and flotation: from stress experiment to stress treatment

    Journal of Environmental Psychology

    (1983)
  • C. Tennessen et al.

    Views to nature: effects on attention

    Journal of Environmental Psychology

    (1995)
  • R.S. Ulrich et al.

    Stress recovery during exposure to natural and urban environments

    Journal of Environmental Psychology

    (1991)
  • A.E. van den Berg et al.

    Environmental preference and restoration: (how) are they related?

    Journal of Environmental Psychology

    (2003)
  • M.G. Berman et al.

    The cognitive benefits of interacting with nature

    Psychological Science

    (2008)
  • Bird, W. (2007) Natural thinking: investigating the links between the natural environment, biodiversity and mental...
  • Carrus, G., Scopelliti, M., Bonaiuto, M., & Romoli, E. (2008). On-site evaluation of restorative environments: the...
  • P. Dolan et al.

    How can measures of subjective well-being be used to inform public policy?

    Perspectives on Psychological Science

    (2007)
  • Cited by (0)

    View full text