For other segments of the Science of Scenery, click here:

TYPOLOGIES OF LANDSCAPE STUDIES

Making sense of the vast variety of landscape studies over the past 35 years or so has been a continuing focus of many researchers in the field. Typologies or classifications of landscape studies are reviewed here. An alternative typology is also presented.

Click on the following:

Description of typologies
Evaluation of typologies
An alternative typology

Description of Typologies

Penning-Rowsell, 1973

 In 1973, Edmund Penning-Rowsell carried out an early review and separated the studies into two types: those independent of landscape users, and those dependent on landscape users. Most of the studies were of the first type, and these generally involved the user defining their preferences rather than the researchers observing the users’ exhibited preferences.

Brush, 1976

In 1976, Robert Brush distinguished between two types of observer-based assessments:

Preferential judgement   This approach elicits an individual's like or dislike for a specific environment. The wide range of personal biases, tastes, and inclinations reflected in preferential judgement is likely to result in a wide variation in responses. This is hardly a firm basis for establishing standards of perceived environmental quality.

Comparative appraisal  This approach forces the observer to adopt a certain framework for making judgements, a context that compels the observer to consider the expected appraisal of a larger group of persons. If an observer were forced to adopt a certain psychological set, their assessment would reflect the values that they ascribe to a larger group. The variation in responses of several individuals would be less than the variation in preferential judgements of the same individuals. Therefore, this latter assessment may be more useful in public decision-making.

Dearden, 1977

A further early typology was an annotated bibliography of landscape aesthetics by Phillip Dearden [1977], in a Council of Planning Librarians Exchange Bibliography. Dearden identified two main groupings of studies:

• Measurement techniques, in which physical attributes of the landscape are used as surrogates for personal perception;
• Preference techniques, in which the landscape is judged in totality, often by reference to criteria established by photographs and questionnaires.

Arthur, Daniel & Boster, 1977

Arthur, Daniel and Boster [1977] presented a synthesis and overview of techniques for evaluating scenic beauty. They grouped the studies into three categories:

• Descriptive inventories: analysis and description of the components of landscapes; e.g. Litton [1968], Leopold [1969]
• Public preference models: assessment based on public input on preferences; e.g. Daniel & Boster [1976]
• Economic analyses: evaluation of nonmarketable environmental goods; e.g. Krutilla and Fisher [1975]

The first two methods can involve quantitative and non-quantitative methods.

Dearden, 1980

 In 1980, Dearden followed up his earlier classification by suggesting three groupings:

• Field-based methods: these involve classifications of the physical landscapes; e.g. Fines [1968], Wallace [1974]
• Surrogate methods: these use observers and photographs as surrogates of the landscape instead of field assessments; e.g. Shafer et al [1969], Daniel and Boster [1976]
• Measurement methods. these refine the measurement quality of field-based methods to improve reliability and validity: e.g. Linton [1968], the Coventry-Solihull-Warwickshire Study [Study Team, 1970].

It is difficult to accept that Linton’s study of Scottish landscapes should be cited alongside the Coventry-Solihull-Warwickshire Study as more rigorous and less subjective than Fines’ study of East Sussex. I include both Fines’ and Linton’s studies in the field-based methods group. Excluding Linton’s study, the measurement method is similar to the psychophysical paradigm [see Zube, et al, below].

 Penning-Rowsell, 1981

Penning-Rowsell [1981] viewed the previous dozen or so years of studies, with a particular emphasis on those in Britain, and divided them into three overlapping groupings:

• Early ‘intuitive’ methods: circa 1967 - 71; e.g. Fines [1968], Linton [1968], Hampshire County Council [1968], Leopold [1969]
• Statistical ‘sophistication’: circa 1971 - 76; e.g. the Coventry-Solihull-Warwickshire study  [Study Team, 1970], the Manchester University evaluation [Robinson, et al, 1976]
• Landscape ‘preference’ approaches: mainly circa 1973 onwards; e.g. the work of Kaplan, Zube and Buhyoff

Penning-Rowsell provided extensive lists of studies in each category. His groupings are somewhat superficial and because they focus largely on Britain [apart from his third group], omit the considerable work undertaken in North America.

 Porteous, 1982

Douglas Porteous [1982] defined four major approaches to environmental aesthetics based on two criteria, rigour and relevance. Porteous noted that, while rigour was traditionally pursued with vigour regardless of relevance, the more recent trend is towards relevance with as much rigour as possible. He defines relevance as referring to the immediacy of the approach to current environmental problems, while rigour refers to scientific theory building and testing [Ibid, 54]. Porteous proposed a model with four groups involved in landscape research [Figure 1].

Figure 1 Porteous’ Groups Involved in Landscape Research

The humanists [or purists] “seek universals intuitively and necessarily eschews immediate relevance and scientific positivism” [Ibid, 54]. Examples are Tuan, Lowenthal and Appleton. The environmental activists seek to ‘act now’ and contrast with the experimentalists who say that ‘before we can change the world, we must first understand it’. Planners is a shorthand term for environmental designers and managers who have to grapple with immediate issues and who often have the training to take a fairly rigorous approach. Porteous considers that no group has reached the “?” position, denoting high levels of both relevance and rigour.

Porteous’ approach tends to diminish the long-term contribution that his so-called humanists [‘theorists’ may be a better term] make. Nevertheless he is correct in identifying two key parameters, relevance and rigour, which should guide work in the field.

 Punter, 1982

A further typology, also published in 1982 though rather lesser known than the others, is that by John Punter. Acknowledging the difficulties in categorising the variety and breadth of the contributions on landscape and the range of disciplines that they derive from, Punter proposes three paradigms, landscape perception, landscape interpretation, and landscape (visual) quality.

• Landscape perception deals with the mechanics of perception and its links with vision, comprehension, preference and action. The roots of this paradigm are psychology and although Punter mentions information theory, he does not refer to the work of the Kaplans.
• Landscape interpretation focuses on the meanings imputed to landscape, especially its social and cultural content. “The comprehension of meaning” according to Punter “involves the search for order and the search for significance.” [Ibid, 105] Yi-Fu Tuan is the leading writer on searching for meaning in landscapes.
• Landscape quality focuses on visual quality and the qualities of formalism apparent in a landscape. Punter considers this the weakest in terms of substantive research yet paradoxically exerting an “alarmingly strong influence” [Ibid, 108] on the experience of landscape.

Porteous considers that Punter “attempts to integrate the three threads via a materialist perspective, a kind of neo-Marxist aesthetic” and that he is “particularly severe on both critics and academic humanists for their ‘privileged indifference’ and detachment” [Porteous, 1996, 12].

Zube, Taylor & Sell, 1982

The two most significant evaluations of landscape studies to date were published in 1982 and 1983. The first was by Ervin Zube, James Sell and Jonathan Taylor, the second was by Terry Daniel and Joanne Vining. Both of these evaluations have a strong orientation to studies from North America. Both identify a range of paradigms into which the various studies were assigned. Interestingly there is a close similarity between the two sets of paradigms.

The Zube et al analysis was based on a review of 160 papers published in 20 journals during the period 1965 - 80 including:

• 63 physical landscape studies
• 21 studies covering recreational activities
• 20 critiques of landscape research
• 16 papers tracing historical or individual [experiential] aspects of landscape

The four paradigms they identified [Ibid, 8] were:

• Expert paradigm - evaluation of landscape quality by skilled and trained observers; skills derive from training in art and design, ecology or resource management
• Psychophysical paradigm - testing general public or selected sample for their evaluation of landscape aesthetic qualities or specific properties. Observer evaluations and behaviour are assumed to bear a correlational or stimulus-response relationship to the external landscape.
• Cognitive paradigm - this involves a search  for human meaning associated with landscapes. Meaning is derived from observation, experience, future expectations and sociocultural conditioning
• Experiential paradigm - the experience of the human - landscape interaction is central here, with both being shaped and shaping by the process.

The paper by Zube et al was a landmark assessment and was accorded an Honour Award by the American Society of Landscape Architects. The Jury stated: “Definitely an outstanding, excellent study. Lots of innovation, marvellous and well-based synthesis. Without doubt one of the leading, current and best-developed documents relating to the field of visual perception and assessment.”  [The Jury, 1982].

As shown in Table 1, the dominant paradigms in Zube et al were the expert and psychophysical. However, whereas the expert paradigm was dominant in the early part of the 16-year period, the psycho physical paradigm became increasingly important as the dominant research direction in the latter part of the period. Zube et al provided a detailed analysis of the studies under each of the paradigms and landscape contexts. On the basis of their findings they recognised the need for the development of a theoretical framework.

Table 1 Frequency of studies per paradigm, 1965 - 80

Characteristic of much of the work has been a separation of two types [Ibid, 23]:

The authors stated: “the most pressing need is for a basic model to which landscape perception research and theory can be fitted and related into a whole.” [Ibid] They presented Figure 2 as a first step towards the development of a theory of landscape perceptions.

Top of next column
Go to top

Source: Zube, Sell & Taylor, 1982
Figure 2 Landscape perception (interaction) process

Zube, et al concluded their paper:

“Research without a general theory is fragmentary and has a hit-or-miss quality to it, it is hard to understand how various research efforts fit together, or indeed, if they are measuring the same thing” [Ibid, 25].

Daniel & Vining, 1983

Following on the heels of Zube et al, in 1983, Daniel and Vining published an independent study with findings that paralleled those in Zube et al. Acknowledging Zube et al, Daniel and Vining termed their paradigms ‘landscape-assessment models’ and defined five such models - ecological, formal aesthetic, psychophysical, psychological, and phenomenological. They described each and evaluated them on the basis of their reliability and sensitivity, validity, and utility.

Both the ecological and formal aesthetic models focus on the characteristics of the landscape whereas the psychophysical, psychological and phenomenological models focus on the effects of the landscape on individuals. Based on their analysis of the reliability, sensitivity, validity, and utility of the models, Daniel and Vining conclude:

“At the present time, none of the models described completely meets all the goals of landscape - quality assessment. By the criteria outlined in this chapter, it is unlikely that either the ecological or the formal aesthetic models can serve as a basis for an adequate landscape - assessment system. For very different reasons, the phenomenological model is inadequate. While neither the psychophysical nor the psychological models are sufficient alone, a careful merger of these two approaches might provide the basis for a reliable, valid, and useful system of landscape - quality assessment.” [Ibid, 80]

Unlike the Zube et al study, Daniel and Vining did not concentrate on the need for theory development. Rather their emphasis was on the improvement of models of landscape analysis.

Go to top

Comparison of Zube et al and Daniel & Vining

As stated earlier, there are close parallels between the Zube et al and Daniel & Vining classifications [Figure 3]. Although Daniel and Vining’s ecological model is based on expert opinion, it also reflects a strong naturalism ethic and defines landscape quality in biological rather than aesthetic terms.

Figure 3  Comparison of Landscape Typologies of Zube et al and Daniel & Vining

The expert paradigm and formal aesthetic model involve assessments of landscapes in terms of their abstract features, including lines, forms, colours and textures, by persons skilled in making such judgements. The psychophysical paradigm/model establishes quantitative relationships between physical features and human responses through testing of observers’ preferences. The cognitive paradigm/ psychological model focuses on the feelings and perceptions of people who interact with the landscape and the meaning that land scapes can hold for people. The experiential paradigm/ phenomenological model focuses on the individual experience of the human-landscape interaction, a person’s subjective feelings, expectations and interpretations in an encounter with the landscape.

Since these two seminal works, further systems for classifying the growing landscape literature have been proposed.

Fenton & Reser, 1988

Fenton and Reser [1988] classified the approaches into three categories:

• Objective measurement of physical-setting variables
• Use of judges’ ratings (normative judgements) to define landscape variables with a clear environmental referent
• Description of landscape variables in phenomenological terms

Their first category combines aspects of psycho-physical and expert paradigms, while the second category covers the cognitive, psychophysical and expert paradigms and the third category covers the experiential paradigm. Fenton and Reser suggest an integrative approach to the perception of landscape quality, in which the “inter-relationships between the objective and the perceived dimensions of the environment are simultaneously examined in terms of their contribution to perceived aesthetic quality” [Ibid, 117]. Essentially, they propose relating the landscape’s physical attributes with the corresponding judgements and deriving the relationship between the two. This is essentially the methodology of psychophysical studies.

Dearden & Sadler, 1989

Dearden and Sadler [1989] developed a theoretical framework based on whether the landscape judgement is a mixture of elements external to the observer [i.e. objects] or internal to the observer [i.e. the perceptual, affective and cognitive responses] [Figure 4].

Source: Dearden & Sadler, 1989
Figure 4 Theoretical Framework based on Consensus for Landscape Evaluation

The ratio of external [E] and internal [I] elements varies with the characteristics of the observer, the landscape and the mode of interaction. Where E exceeds I [E > I], consensus will be high, but where I exceeds E [I > E], consensus will be low. E > I is termed objectivist, while I > E is termed subjectivist. The authors compared their framework with the five models defined by Daniel and Vining [1983].

While they acknowledge that it is often difficult to assess the I:E ratio, Dearden and Sadler considered that “some techniques, firmly rooted in an objectivist philosophy, are purely landscape oriented and merely assume consensus”,  whereas “other techniques pay little attention to landscape, assume that each observer is unique, [that] there is no consensus and focus their efforts on a subjective analysis of the individual.” [Ibid, 9] On the basis of their analysis, the authors suggest that the various approaches to assess landscape quality “should not be seen as mutually exclusive, ... [but] rather they are complementary”.

Elsewhere, Dearden [1989, 41] defined the objectivist stand of viewing beauty inherent in objects, whereas the subjectivist stand views beauty as being in the eye of the beholder.

At one extreme, the ecological model is focussed wholly on the landscape with no human input while at the other extreme, the phenomenological model is wholly focussed on the human experience and the landscape is almost incidental. The models lying between these two extremes contain elements of both landscape and human involvement. The psychophysical model straddles each and aims to measure both the landscape and human response to it. Dearden and Sadler consider that the psychophysical approach is probably the most favoured approach with good reliability and validity. The phenomenological approach is regarded as the most ‘scholarly’, while practitioners favour the formal aesthetic.

Dearden and Sadler’s identification of the objectivist and subjectivist elements in landscape assessments is welcome, although they appear to confuse objectivity with consensus. Their proposal regarding the relative dominance of external or internal elements appears naïve, as when they state, “in some circumstances beauty will reside more in landscape (i.e. E > I) and in others the eye of the beholder will be more critical in influencing landscape judgements (i.e. I > E).” [Dearden, 1987, 268] This suggests that the influences on an individual are changeable depending on circumstances. Yet it is difficult to see how this could be in practice, how does a person put aside the innate, cultural and personal influences on their preferences and see the landscape purely in terms of intrinsic beauty? The model appears to have the hallmarks of a theoretical construct that has parted company with reality.

Gobster & Chenoweth, 1989

Gobster and Chenoweth defined three ‘descriptor types’: physical, artistic and psychological and analysed these types’ capacity to predict aesthetic preferences for rural river, forest and agricultural landscapes.

Evaluation of typologies

Table 2 summarises the ten typologies that have sought to classify the landscape studies. It lists them under the four paradigms defined by Zube, et al, 1982 for comparison. In some cases it is difficult to assign the typologies as the descriptions used differ greatly, however the Table indicates my best judgement as to their placement. Most of the typologies define only two or three categories and in some instances, several of these amount to the same thing, namely the psychophysical paradigm.

Table 2    Summary of Landscape Analysis Typologies

Note: Arthur et al, 1977 also include economic analyses. The classification by Porteous [1982] is of the type of researcher rather than the product of their work. His activist category does not fit any of the above paradigms. Daniel & Vining [1982] also included the ecological model.

Go to top

An Alternative Typology

The fundamental dichotomy in the way landscape is viewed is between believing that beauty is an intrinsic quality in the landscape versus believing that beauty lies in the eye of the beholder. Surveys of the physical landscape which define its quality on the basis of the presence or absence of certain attributes are premised on the concept of beauty being intrinsic in the landscape. Psychologically-based studies which determine the feelings that people derive from the landscape and which identify the dimensions in the landscape that account for its quality are premised on beauty being in the eye of the beholder.

These different approaches to the way landscape is viewed are quite fundamental, either the landscape quality is regarded as intrinsic or in the beholder - there is no middle ground. Few of the typologies examined recognise this distinction and most treat the differences in the form of a continuum. Where the physical landscape is assessed, it is in terms such as field based, descriptive inventories, expert, or objective measurement. Gobster & Chenoweth (1989) touch on the differences in stating: “All physical descriptors relate to the external dimensions of the environment - what is ‘out there’ versus what is ‘in the head’ - and herein lies a critical distinction between physical and psychological descriptors.” Similarly Dearden and Sadler (1989) come close to the issue in stating:

“The major philosophical and methodological division has been between those favouring a more reductionist, quantitative-objective approach and those maintaining that it is not possible to apply standard positivist techniques to such a holistic concept as landscape aesthetics.”

I propose that this distinction should provide the basis for the major classification of landscape methods, between physically based methods and psychologically based methods, the former being those based on viewing beauty as physically intrinsic in the landscape while the latter view it as a human preference. I term these the physical and the preference paradigms.

Dearden and Sadler (1989) proposed a similar objectivist/subjectivist terminology depending on whether the landscape judgement is based on elements which are, respectively, external or internal to the observer, ie:

Table 3 summarises the differences between these two paradigms. Based on the differentiation between the physical and the preference paradigms, the various models and methods proposed by various writers could be assigned in an hierarchical manner.

Table 3 Physical and Preference Paradigms

Although this framework groups the various paradigms and models of authors together, it is recognised that there are substantial differences between each of them, e.g. between the cognitive and phenomenological paradigms. However the emphasis in this framework is to differentiate between the physical and the preference paradigms, the differences within each are of much less importance than this fundamental distinction.

Table 4 illustrates the placement of the paradigms by Zube et al and Daniel & Vining.

Table 4 Hierarchy of Landscape Assessment Methodologies

Go to top