Richard Neutra: Spatial Theory and Practice

Chapter
Part of the Mathematics and the Built Environment book series (MABE, volume 3)

Abstract

While Richard Neutra is conventionally celebrated as the archetypal Modernist architect, his designs were only superficially indebted to the tenets of European Functionalism and the aesthetic values of the International Style. He was instead profoundly influenced by scientific theories that sought to measure and predict the way the human body would react to space and form. These theories led him to design buildings in such a way as to choreograph people’s emotional and physical responses through a process called ‘visual excitation’. For Neutra, visual excitation is triggered by controlling the way people see, move through and comprehend space.

While Richard Neutra is conventionally celebrated as the archetypal Modernist architect, his designs were only superficially indebted to the tenets of European Functionalism and the aesthetic values of the International Style. He was instead profoundly influenced by scientific theories that sought to measure and predict the way the human body would react to space and form. These theories led him to design buildings in such a way as to choreograph people’s emotional and physical responses through a process called ‘visual excitation’. For Neutra, visual excitation is triggered by controlling the way people see, move through and comprehend space.

Neutra’s (1962, 1971) argument is founded in two major beliefs; that vision leads to movement and experience leads to understanding. In practical terms, the first position maintains that long, controlled vistas in a design stimulate a person’s visual senses, drawing them through the building and making them aware of its larger context. The second belief holds that this experience of movement provides each person with a strong sense of where they are in space (within both the building and its larger context) and in time (being aware of diurnal and seasonal changes). This last property has also been linked to an additional proposition from Neutra, concerning the ontological importance of the direct experience of nature.

This chapter uses syntactical analysis to investigate whether Neutra designed in accordance with these three theories about the relationship between vision and movement, cognition and experience, and interior and exterior. Focussing on Neutra’s Californian houses, the chapter uses axial line analysis to search for evidence of this connection between spatial theory and practice.

6.1 Introduction

Austrian-born American architect Richard Joseph Neutra produced designs that are typically interpreted as using technology and industrial materials to express their functional properties and the spirit of their era (Boesiger 1964; Hines 1982; Sack 1992). He has been described as one of ‘the most celebrated of the founders of Modern architecture’ and as a designer ‘who managed to capture the spirit of Modernism in a powerful and memorable way’ (Sennott 2004: 917). However, despite his reputation for producing white, geometric designs, which otherwise appear to conform to early twentieth century Modernist ideals, Neutra repeatedly claimed that his primary purpose as a designer was to shape the sensory and cognitive responses of the human body (Neutra 1962, 1971).

Long before other architects were attracted to phenomenological reasoning, Neutra argued that architecture’s raison d’être is to provide a type of sensory and physical framing to assist each person to understand and appreciate their position in the world (Neutra 1951, 1956). While his position has parallels with those of later phenomenological thinkers in architecture, it also departs from the then more recent tradition privileging science over philosophy. Fundamentally, the intellectual lineage of the majority of contemporary architects who are working in the phenomenological tradition can be traced to philosophers Edmund Husserl and Martin Heidegger, whereas the origins of Neutra’s theory are found in the work of experimental psychologist Wilhelm Maximilian Wundt.

Both the scientific and philosophical traditions of twentieth century thinking about the senses have their origins in Germany in the late nineteenth century. At that time Wundt began formulating a way of isolating and measuring bodily reactions to external stimuli, including heat, sound or smell. Wundt believed that by understanding the senses and their impact on behaviour, human responses could be modified or even, for clinical purposes, controlled. Husserl (1982) strongly rejected this approach, calling for the primacy and irreducibility of human experience to be recognised. Heidegger (1962) developed this argument to support an inquiry into the nature of being and existence, and in the hands of late twentieth century architects (including Peter Zumthor, Steven Holl and Juhani Pallasmaa), the spiritual dimension of Heidegger’s philosophy was translated into an argument for the transcendent quality of certain places and buildings (Norberg-Schulz 1979). These architects maintain that great design evokes a deep, sensory appreciation of place, space or tectonics and that it is only through the search for authenticity and truth that this can be achieved (Pérez-Gómez 1983; Pallasmaa 1996).

The division between contemporary architectural phenomenology and Neutra’s theory of sensory choreography is essentially one between, respectively, poetry and science. This is not to say that Neutra’s ideas lack poetry, nor that the architectural phenomenologists reject science. Instead, Neutra’s ideas emphasise a particular empirical mindset and a clinical foundation for his design theory, while those of the architectural phenomenologists tend to valorise the metaphysical. In much the same way that Wundt and Husserl were both fascinated with the role played by the human senses in understanding the world, but resorted to different ways of achieving their ontological goals, so too Neutra cannot be so easily detached from the more recent phenomenological tradition.

These characteristics of Neutra’s architecture begin to explain the contradictory place he occupies in the history of Modernism (Hines 1982; Lamprecht 2000). On the one hand, his work appears to embody all of the values of the Modern movement, but the theories he used to explain his designs do not repeat the classic Modernist tropes of functionalism and the zeitgeist. Further exacerbating this problem, Neutra’s theory has rightly been described as structurally ‘unsystematic’ and his prose as ‘hard to read’ and ‘repetitive’ (Kruft 1994: 432). Thus, while Neutra offered lengthy, detailed descriptions of his design process, their complexity and opacity led many historians and critics to ignore his theories altogether and interpret his architecture as a type of socially-informed, technologically-enabled response to the era (McCoy 1960; Boesiger 1964).

It has only been in the last few decades that researchers have begun to take Neutra’s claims about his design process and representational practices more seriously (Lavin 1999, 2000, 2004; Lamprecht 2000; Ostwald 2014b). For example, Neutra preferred to describe his architecture using a combination of perspective views and plans—his dislike of elevations setting him apart from his Modernist contemporaries. But where this attitude was once dismissed as a symptom of Neutra’s obsession with the image, more recently scholars have asked whether his rejection of the elevation might be integral to the way he visualised people’s experience of space (Niedenthal 1993; Lamprecht 2000; Ostwald and Henderson 2012). Collectively, the resurgence of interest in Neutra is partly associated with a simple question, is there evidence that he applied his theory in his architecture?

This question is the catalyst for the present chapter, which proposes a syntactical analysis of the planning of five of Neutra’s Californian houses: the canonical Kaufmann Desert House from 1947, and the Tremaine, Moore, Kramer and Oxley houses from, respectively, 1948, 1952, 1953 and 1958. The purpose of this research is to examine three related characteristics of Neutra’s design theory: the use of long sight-lines to control the way movement occurs; the use of visual and movement patterns to support or enhance spatial cognition; and the role of the exterior in the social and experiential function of the home.

For the first of the three characteristics, which is associated with Neutra’s argument that vision leads to movement, axial line analysis is used to examine the topological significance of sight lines and paths in Neutra’s architecture. Combining measured values (depth and integration) alongside a review of the spaces surveyed and passed through as part of the structure of the house (its non-trivial loops and hierarchies), the chapter examines whether long sight and movement lines are especially significant in these plans. For the second characteristic being analysed, which arises from Neutra’s claim that experience leads to understanding, a correlation is constructed between integration and connectivity measures to calculate the relative intelligibility of each plan. This process identifies the degree to which paths and vistas in the house allow a visitor to efficiently experience and thereby comprehend a design.

The third characteristic of Neutra’s theory examined in this chapter is also associated with the argument that experience leads to understanding, because Neutra maintained that it is critical for people to appreciate not only their domestic environment, but also the context in which it is situated. In practical terms, Neutra proposed that people should experience the passing seasons, and changes in temperature and weather, using all of their senses (Neutra 1956, 1962). He had two strategies for achieving this outcome. First, people in the interior of a house, regardless of the social function of the space they are inhabiting, should not only be able to see outside, but move outside with relative ease. Second, people should be encouraged (or even required) to move outside as part of their quotidian occupation of the house. In this chapter, the first of these is examined by determining the step distance of each plan, that is the number of turns required to navigate the ‘shortest path from the formal entrance (most public space) to the main bedroom (most private space) of each dwelling’ (Hanson 1998: 243). For the second, the extent to which exterior connections are necessary for the social functions of the house is determined through a qualitative review of the axial maps.

Importantly, this chapter does not test if Neutra’s theory of visual excitation actually works. That is, this chapter does not examine whether people’s emotional and physical responses can be choreographed in any methodical or consistent manner by architecture. While this proposition is inherently seductive for many architects, there is little or no empirical evidence that supports the validity of determinism of this kind. Certainly some behavioural and environmental preference research suggests that people are statistically likely to experience mild positive emotional responses to certain spatial, formal or symbolic stimuli (Stamps 2000). For example, people do tend to prefer views of natural settings over views of urban settings or spaces with no outlook at all. People also tend to be happier if they have access to natural light and ventilation (see Chaps.  8 and  9 for a more detailed discussion of this issue). Such findings have been used to support the ‘biophilia hypothesis’ which claims that humans have an innate affinity to natural forms and environments (Kellert and Wilson 1993). This idea would have resonated with Neutra, who called his own theory of architecture ‘Biorealism’, and who used views of nature, typically framed and controlled by architectural form, as a means of assisting his clients to locate themselves in space and time. However, the efficacy of Neutra’s theory is not the subject of this chapter, which is instead concerned with investigating whether the spatial strategies espoused by Neutra are actually present in his designs.

6.2 Neutra and Biorealism

In his 1956 book Life and Human Habitat, Neutra outlined his theory of Biorealism, which called for architecture to strive to achieve three interconnected goals. The first is to support the human body to reconnect with nature; the second is to limit the impact of chaotic environments that would otherwise distract the mind. The third is to address the sensory needs of the human body in such a way that it can better understand the environment. While the underlying anthropological rationale for this theory is beyond the scope of this chapter, what is more interesting in the present context is how Neutra set about achieving this outcome in his architecture.

The essence of Neutra’s design strategy can be traced to the results of Wundt’s nineteenth century experimental research, which uncovered a series of relationships between human perceptions and reactions. In particular, Wundt observed that ‘the reflex to the muscles that move the eye-ball is connected … with contraction of the corresponding muscles for movement of the head’ (1969: 294). This finding suggests that vision (as translated through the musculature surrounding the eye) has a direct connection to bodily responses (the movement of the head). Neutra was fascinated by this idea and extrapolated it to claim that if the eye is drawn to a particular vista, then the head will turn towards that view, which in turn will be the catalyst for the entire body to move in that direction. This led to the formulation of Neutra’s famous maxim; ‘we “see not merely to see” but see in order to act upon vision’ (Neutra 1956: 13). Vision, he argues, activates ‘a person’s locomotor urges’ (Neutra 1956: 14), causing them to respond physically, and through this movement, to learn about the spaces and environments they are inhabiting. Furthermore, if this relationship between vision and action, or sense and response, is universal, then architectural form (what Neutra called ‘shape’) can be used to trigger and control these urges, causing a person to look in a certain direction and follow a particular path. Neutra summarised this idea as being that human response is a ‘consequence of shape’ (1956: 20).

Neutra developed and presented this theory about sensory response and its choreography through design, using his residential works of the late 1940s and early 1950s as examples. He did this because he believed that the home is the single place on ‘the surface of the globe which we get to know intimately’ (Neutra 1956: 21). Yet the home is also, paradoxically, the site of much of the ‘optical litter’ and ‘visual conflict’ that prevents the body connecting with and appreciating its environment (Neutra 1956: 166). The architect’s role is therefore to minimise the chaotic or random distractions of the home, using a limited palette of forms (flat roofs, orthogonal walls and planes of glass) to shape the way the eye is captured, the body is inspired to move, and the mind is led to understand its environment.

While Neutra repeated this argument throughout much of his career, the biggest impediment to accepting it as a true account of his practical approach is that he rarely provided any direct evidence of how he applied it in his designs. At best, there are multiple clues in Neutra’s books that reveal how he might have designed to achieve these outcomes. For example, his tendency to imagine and depict spatial affects using perspectives sketches, sometimes keyed to specific locations in a plan, reinforces the notion that he designed to achieve particular visual effects. He also went to great pains to minimise visual distractions in some parts of his design, but not in others, which suggests a deliberate strategy at work. For example, he sometimes minimised the division between the interior and the surrounding landscape by using corner glazing (without mullions) and emphasises the connection using continuous ceiling-eaves and reflecting pools. But more importantly, he occasionally used silver paint on structurally indispensable elements in an attempt to ‘dematerialise’ them, so as not to distract the eye from the vistas and paths he wished to emphasise. As Lamprecht observes, the silver finish ‘appears only in those places where the sightline is affected’ (2000: 34). Thus, for some vistas, there was a conscious decision to minimise distractions, while for others this was deemed relatively unimportant.

Another example of how Neutra worked is found in his various imagined accounts of people experiencing, reacting to and understanding architecture (Ostwald and Henderson 2012). For example, as mentioned in Chap.  1, he describes the experience of a person approaching a house as follows. ‘As we approach we raise our head to recognize the house number [and as] we tilt our head upward, the equilibrium or inner ear organ immediately functions and combines the manifold record of our body position with pure vision and its ever-changing perspectives’ (Neutra 1956: 13). This quote is notable for a number of reasons, not the least of which is the way it has been constructed using the first-person, collective pronoun ‘we’. As Neutra’s account progresses, the clear inference is that a universal perspective is being presented and that the human eye—indeed any human eye—is inexorably drawn to the house number, then the angle of the head is inclined to take in the entire house and its natural context. Neutra’s description continues in this vein, stating that we then ‘roll our eyes by means of that ingenious muscle cluster around our eyeballs which is intricately and neurally tied up with those tools which we use unconsciously for turning and tilting the head’ (Neutra 1956: 13). Here, once again, Neutra stresses the involuntary response of the body to stimuli in general, and architectural form in particular. Before we know it, we have ‘our hand touching the knob of the entrance door, tactile and thermal experience of conductive and polished metal comes to us through the fingers and palms of the hands, while at the same time the muscle senses faithfully report from below about the rubber mat on which we have stepped’ (Neutra 1956: 13). In this example, architectural form stimulates the senses in such a way as to lead the body from the street to the front door, providing the mind with a subconscious understanding of the spatial approach to the house. Furthermore, while we have remarked on the first-person framing of Neutra’s narrative, its present tense is also notable because it emphasises the immediacy of the reactions; they are subconscious responses to the controlled use of architectural form.

Ultimately, the success of Neutra’s theory of Biorealism rests on his capacity to achieve three properties in a design. First, the creation of long, distinct vistas, to draw the eye and the body through space. Second, the creation of plans with improved connectivity to support increased cognitive clarity. Third, to ensure that the exterior environment is part of the experience of the house. Through the application of these strategies, his architecture was intended to promote heightened sensory appeal, reduce distractions and increase awareness of nature. These strategies are the subject of the remainder of the chapter, insofar as evidence of their application can or cannot be found in five of Neutra’s most important designs of the era.

6.3 Method

6.3.1 Hypotheses

Without a mathematical basis to directly compare Neutra’s houses with other designs, the three hypothesises which drive this chapter are formulated more to guide the discussion and conclusion than as definitive results (Table 6.1). In all cases, the test is framed as evidence being found in at least four of the five cases, because these are all houses that Neutra used to support his argument, and so a greater than average number of positive results should be anticipated. The results for two of the three hypotheses also combine quantitative and qualitative indicators.
Table 6.1

Spatial properties mapped to specific hypotheses, analytical methods and result indicators

 

Property

Hypothesis

Method

Indicator of a positive result

1

Vision leads to movement

Long sight lines will dominate the network of vistas and paths in each plan

Axial line analysis

The most integrated line(s) should encompass a large proportion of each plan’s functional zones and participate in their major circulation loops

2

Movement leads to understanding

Plans will possess a high level of cognitive clarity

Intelligibility comparison

R2 > 0.75 in at least four cases

3

Experience of the exterior is integral

(i) The most private spaces will be topologically close to the exterior

(ii) The structure of the plan will require exterior connections

Step distance and structural comparison

(i) Step distance will be less than 2 in at least four cases

(ii) Exterior spaces will be integral to the social structure of at least four cases

To assess the first hypothesis, Total Depth (TD) and Mean Depth (MD) are employed to classify and differentiate those spaces that are shallow or deep relative to the entire building, and integration (i) is calculated to investigate how accessible a line is to every other line in the system. Integration was calculated from Real Relative Asymmetry (RRA), which means that the results are relativised for direct comparison between the differently sized houses (Turner 2004).

For the second hypothesis, Intelligibility (I) is used as the mathematical basis to construct a comparison (see Chap.  2). Intelligibility is a measure of how efficiently the configuration of a space can be understood by traversing its component parts (Peponis et al. 1990; Hillier 1996; Haq and Girotto 2003). In early Space Syntax research intelligibility was often represented as a scatter graph of the connection and integration values of each line; if the axes are balanced, then the closer the angle of the line is to 45° the more intelligible the plan. The logic behind this process is that integration represents a global measure of the connectivity of a given space to all other spaces in the system. The number of connections the line makes represents how much of a configuration can be seen from each line and therefore the relationship between these measures suggests how intelligible a plan is. The higher the mathematical correlation of points (R2 → 1) the more intelligible the system. A benchmark for intelligibility values is found in the mean intelligibility of a sample of 75 urban environments: R2 = 0.68 (Hillier et al. 1987). Urban environments with R2 values greater than 0.68 exhibit above average intelligibility. However, for relatively spatially simple environments, like houses, a higher result would be anticipated to indicate above average intelligibility, say R2 > 0.75.

The third hypothesis is tested using step distance, being the minimum number of direction changes encountered along a path between the most public and most private spaces in a design. Hanson (1998) uses step distance as a comparative measure of social separation between visitors and inhabitants, but in our case it serves as the topological distance between inhabitants and the environment.

6.3.2 Approach

As a starting point for the analysis, new CAD models of each of the five houses were prepared using Neutra’s original working drawings along with published plans and photographs of the completed houses. The process used to develop the axial maps from these plans conforms to the general principles described previously (see Chap.  3), where axial lines represent visibility and movement. Not only is the relationship between sight and access pivotal to Neutra’s argument that vision leads to movement, it is ‘one of the most pervasive, effective, and powerful means through which architecture formulates social significance and social meaning is through the separation of accessibility and visibility’ (Koch 2010: 13).

To construct the axial maps, the boundaries of each plan are ‘cropped’ close to the house to maintain the analytical focus on habitable spaces. This means that while exterior paths and driveways in Neutra’s designs can often extend a substantial distance from the house, only those close to the buildings, and clearly ‘spatially defined’, are included. In this case—and because of the importance of outdoor ‘rooms’ and ‘corridors’—‘spatially defined’ is taken to mean any hard-paved area that connects physically separate wings of the design or is required to complete internal circulation loops. Some of these hard-paved areas are also defined by garden or retaining walls, and a few are roofed, providing a stronger sense of these spaces as outdoor rooms.

Spaces that do not contribute to the social function of a building are excluded from the analysis. These include dumbwaiters, plant rooms and storage areas that are either too small for human inhabitation or not designed for any social use. Similarly, built-in furniture is treated as unnavigable obstacles, while all doors are assumed to be open. Following the principle that an axial line represents both vision and access, glazed walls are treated as obstructions to movement, as are pools and water features.

Stairs in the five houses are divided into two categories: those which do not interrupt the spatial cognition of a user and those that do. In the first category are small sets of stairs, with four or fewer risers, that are part of a larger space and do not unduly disrupt the direct visual and physical connection across that space. In the second category are stairs with five or more risers, which are treated as separate spatial units requiring dedicated axial lines. Landings are considered part of any stair set connected to them; thus a single landing could be simultaneously considered as part of both types of stair under the additional selection logic applied. Where a dedicated line is required to connect multiple levels, this section of the plan is split as necessary. Where no axial lines exist that run the entire length of a stair, one is added to complete the circulation paths and a line connecting it to the remainder of the network is also included. These lines are annotated (‘A’ and ‘B’) on the axial plans of the Kaufmann and Tremaine houses, and an arrow included demonstrating the connection.

For each house a perspective view is provided along with an annotated plan identifying the original functional uses of the spaces (Table 6.2). The first numbered bedroom in each plan is the master bedroom. Maid’s and chauffeur’s bedrooms, guest bedrooms and servant spaces (including small living rooms for servants) are differentiated from the main family spaces in the text, but not in the annotations, which are simply numbered in accordance with their base function.
Table 6.2

Key to plan annotations

Abbreviation

Room

C

Courtyard

CP

Car parking

T

Terrace

GL

Gloriette

E

Entry

H

Hallway

P

Porch

K

Kitchen

L

Living

D

Dining

G

Gallery

ST

Studio

LB

Library

B1

Master bedroom

B2, B3, …

Secondary bedrooms

b1, b2, …

Bathrooms

DR

Dressing

S

Storage

LD

Laundry

SV

Service area

CR

Change rooms

For the axial map of each house, the lines are differentiated to show the top third (most integrated), the middle third, and the lowest third (least integrated). The most integrated line is numbered, along with the line that is closest to the mean and the one that is least integrated in the plan (Table 6.3). If two or more of these lines have equal i values, then they are all signified in this way. Rather than providing a complete table of results for every line in each house, only the top three, middle three and lowest three lines, based on integration values, are inserted in a table. However, because multiple lines may have identical i values, several of the tables contain more than these nine results. The high, mean and low results, at the base of each column in the tables, are for the complete set of lines. Each table also includes the total number of lines in the map and the step distance between the master bedroom and the entry space in the house.
Table 6.3

Key to axial maps

Category

Type

Key

High integration

Most integrated line(s)

Open image in new window

Top third of integrated lines

Open image in new window

Middle integration

Mean integrated line(s)

Open image in new window

Middle third of integrated lines

Open image in new window

Low integration

Least integrated line(s)

Open image in new window

Lowest third of integrated lines

Open image in new window

6.4 Results

6.4.1 Kaufmann Desert House, Palm Springs, California, USA (1947)

The Kaufmann Desert House is sited at the base of Mount San Jacinto in Palm Springs, California. Described by Barbara Lamprecht as both Palm Springs’ ‘first Modernist grand villa’ and a ‘social extrovert’ (2000: 179), the house ‘subsequently became the chief d’oeuvre [of the] suburban townscape’ (Hines 1982: 201). The plan of the Kaufmann Desert House features a central living area from which four major functional wings radiate. This pinwheel configuration situates more public spaces closer to the core and more private or functional zones (such as the garage) to the wings (Fig. 6.1). Significantly, the rooms in three of the wings are only accessible by way of external circulation paths. The Kaufmann Desert House also possesses two prominent orthogonal axes that bisect the plan: the first links the garage and entry to the guest rooms and the second links the maid’s room and the bedrooms. Finally, centred over the living area is an upper level gloriette (from the French word gloire, meaning ‘little room’), a semi-external living space (Fig. 6.2). The reliance on external paths and spaces for its basic functional operation is a feature of this design.
Fig. 6.1

Kaufmann Desert House, perspective view

Fig. 6.2

Kaufmann Desert House, annotated plan

The axial map for the Kaufmann House identifies thirty-three lines, the longest and most integrated of which cross in the living room (Fig. 6.3, Table 6.4). The most integrated line (i = 3.27) runs from the second maid’s bedroom (B5), along an exterior corridor, through the entry foyer and the living room to the pool. The line with the integration value closest to the mean commences in the first maid’s bedroom (B4) before passing along an exterior corridor to the dining room. The gloriette—which is connected to the house by way of a single set of external stairs that are accessed from the pool area—contains the most segregated line in the design (i = 0.68). The complete set of results confirms that the Kaufmann Desert House is flexible and adaptive when external circulation connections are included and highly inflexible, and indeed non-functional, when they are excluded.
Fig. 6.3

Kaufmann Desert House, axial map

Table 6.4

Kaufmann Desert House, results

 

MD

RA

RRA

i

Line length

Connectivity

Lowest three i values

4.9062

0.2520

1.4712

0.6797

13,217.8010

1

3.9375

0.1895

1.1063

0.9038

11468.7530

2

3.8750

0.1854

1.0828

0.9235

4547.7354

1

Median three i values

2.8750

0.1209

0.7061

1.4160

5721.0405

3

2.8750

0.1209

0.7061

1.4160

9628.5596

3

2.8750

0.1209

0.7061

1.4160

6052.4800

4

2.7500

0.1129

0.6591

1.5171

7499.2041

2

2.7500

0.1129

0.6591

1.5171

8289.5205

2

Highest three i values

2.0625

0.0685

0.4001

2.4988

21077.9380

10

2.0312

0.0665

0.3884

2.5746

21867.7210

9

1.8125

0.0524

0.3060

3.2677

38547.8590

14

Min

1.8125

0.0524

0.3060

0.6797

4081.13960

1

Mean

2.8446

0.1190

0.6947

1.6279

13669.5777

4.4242

Max

4.9062

0.2520

1.4712

3.2677

38547.8590

14

Step distance (E → B1)

2

Total lines in map

33

Only two changes of direction are required to progress from the formal entrance to the master bedroom, a result which is also the same for three of the four remaining bedrooms. The second maid’s bedroom requires only a single change in direction from the entry. This means that each of the deeper bedrooms is at least partially visible with only a single direction change after entering the house. This implies either a relatively low privacy threshold in the house or a high degree of connection to the exterior. The data also indicates that the Kaufmann Desert House is very intelligible or has a high level of cognitive clarity about it: R2 = 0.8764 (Fig. 6.4). This result occurs for two reasons: there are several highly integrated, long lines that pass through the centre of the plan, and the overall planning strategy for the house is relatively shallow (MD = 2.8).
Fig. 6.4

Kaufmann Desert House, intelligibility graph

6.4.2 Tremaine House, Montecito, California, USA (1948)

While Neutra was working on the Kaufmann Desert House , he was also designing the Tremaine House , and the two share several features. For example, both have pinwheel plans, centrally located social spaces and masonry walls grounding their structures, visually at least, in the landscape (Fig. 6.5). The Tremaine House was designed as a permanent residence for a couple and their three children. It has more than thirty rooms, including a basement level (Fig. 6.6). As is the case with the Kaufmann Desert House, in the Tremaine House a small number of long vistas and associated paths bisect the centre of the elongated pinwheel plan, forming connections between the different wings and many of the spaces in between. However, in the Tremaine House there are non-trivial circulation loops which do not require external circulation, offering a greater degree of flexibility in the way the space is inhabited and used at any time of the year and under any weather conditions.
Fig. 6.5

Tremaine House, perspective view

Fig. 6.6

Tremaine House, annotated plan

The axial analysis of the Tremaine House reveals that fifty-six lines are required to produce a valid map (Fig. 6.7, Table 6.5). The data from the map identifies that the longer lines crossing the central living spaces are the most integrated (2.11 < i < 2.30) and the third highest integration value is for a line crossing the formal entrance. Conversely, the lines in the basement level are amongst the least integrated in the design (0.58 < i < 1.0). The line with the highest integration value passes through the terrace, the living room, the kitchen and the exterior circulation. The lines that are closest to the mean for integration pass through, respectively, the hall and bedroom 2, and the library and the filing room. The lines with the lowest integration values are found in the change rooms in the basement level (which were intended for a pool that was never built).
Fig. 6.7

Tremaine House, axial map

Table 6.5

Tremaine House, results

 

MD

RA

RRA

i

Line length

Connectivity

Lowest three i values

6.7090

0.2114

1.7038

0.5869

1992.0631

2

6.7090

0.2114

1.7038

0.5869

1993.1252

2

5.7636

0.1764

1.4216

0.7033

5314.2666

4

5.7636

0.1764

1.4216

0.7033

5314.2012

4

Median three i values

3.5090

0.0929

0.7488

1.3354

23184.8200

5

3.4727

0.0915

0.7379

1.3550

9551.1729

3

3.4727

0.0915

0.7379

1.3550

6084.0195

4

Highest three i values

2.5818

0.0585

0.4720

2.1182

33713.6410

18

2.5272

0.0565

0.4558

2.1939

24324.9860

12

2.4545

0.0538

0.4341

2.3036

40290.1450

16

Min

2.4545

0.0538

1.7038

0.5869

1992.0631

1

Mean

3.7409

0.1015

0.8180

1.3597

11828.1753

5.3214

Max

6.7090

0.2114

0.4341

2.3036

41127.5200

18

Step distance (E → B1)

1

Total lines in map

56

A straight line of sight from the entry looks across the living spaces into the vestibule of the master bedroom and, with a single turn, the bedroom itself is visible, meaning that this plan has a step distance of one. Despite this seemingly shallow social depth, the Tremaine House is actually the least intelligible of the five Neutra houses analysed in this chapter (R2 = 0.54921), having both the weakest correlation between line connectivity and integration and the least balanced (or furthest from 45°) trend-line where the axis is scaled to give a square graph (Fig. 6.8).
Fig. 6.8

Tremaine House, intelligibility graph

6.4.3 Moore House, Ojai California, USA (1952)

The Moore House deviates from Neutra’s typical pinwheel planning arrangement of the era, having a more linear, dual pavilion parti, with all of its communal spaces to the north (Fig. 6.9). Like the Kaufmann Desert House , the guest wing and parking areas are accessible only by way of external circulation routes. A corridor connects the communal, private and service spaces and is extended to become the path to the guest wing. The linearity of the plan led Sylvia Lavin to observe that experientially, ‘the architecture and the views it offers … are captured through peripheral vision’ (2004: 109). With twenty-three defined spaces, the Moore House is also much smaller than the Tremaine House and about the same size as the Kaufmann Desert House (Fig. 6.10).
Fig. 6.9

Moore House, perspective view

Fig. 6.10

Moore House, annotated plan

The axial map for the Moore House reveals that two long perpendicular axes form the central connections within the main house (Fig. 6.11). In this sense, the Moore House has a similar visual structure to the previous two, despite not having the same planning strategy. The map also reveals that three non-trivial internal circulation loops exist, including, surprisingly, two involving the master bedroom and bathrooms 3 and 4. The third loop encompasses all the house’s major living areas. This property of the plan suggests that the owners have been provided with a much higher degree of flexibility than their guests and, like the Tremaine House, the master bedroom is directly accessible from the formal entrance. Only two circulation loops which involve external connections exist, both of which include the hallway.
Fig. 6.11

Moore House, axial map

The most integrated line in the map passes from the exterior circulation area, through the hall, to the service area and the kitchen (Table 6.6). This path effectively functions as the service backbone of the house; this strategy differentiates this design from the previous two, in which the most highly integrated lines were associated with living spaces. The mean integrated line commences in the master bedroom (B1) and progresses through the bathroom (b4) and into the hall. The least integrated vistas and paths are those from the studio to the patio, and from the patio to the guest bedroom (B3). Curiously, the formal entrance is on the third most integrated line. The most and second most integrated lines intersect at a shallow angle above the upper level external stair. Bedrooms in the main house are of approximately average integration (0.79 < i < 1.20) and the living areas fall into a similar range. Like the Tremaine House , this design requires only a single direction change between the formal entrance and the master bedroom (B1). This result supports the general reading of the plan as containing a strangely centrally located master bedroom, with three alternative entry paths into it.
Table 6.6

Moore House, results

 

MD

RA

RRA

i

Line length

Connectivity

Lowest three i values

6.1923

0.4153

2.1644

0.4620

8102.5996

1

6.1923

0.4153

2.1644

0.4620

7977.4438

1

5.2307

0.3384

1.7636

0.5670

11600.3780

3

5.2307

0.3384

1.7636

0.5670

8147.4399

3

Median three i values

3.3461

0.1876

0.9780

1.0224

8152.3281

3

3.2307

0.1784

0.9299

1.0753

15731.9580

5

3.2307

0.1784

0.9299

1.0753

14989.4880

4

Highest three i values

2.6153

0.1292

0.6733

1.4850

30626.2990

10

2.5384

0.1230

0.6413

1.5592

22358.2300

7

2.3461

0.1076

0.5611

1.7820

26545.8610

9

Min

2.3461

0.1076

0.5611

0.4620

2964.9650

1

Mean

3.6011

0.2080

1.0843

1.0436

12294.4870

4

Max

6.1923

0.4153

2.1644

1.7820

30626.3000

10

Step distance (E → B1)

1

Total lines in map

27

The intelligibility graph for the Moore House has an R2 value of 0.6783, which confirms a moderate to low level of correlation between integration and connectivity (Fig. 6.12). The angle of the balanced trend-line also supports the idea that the Moore House is at best modestly intelligible. However, despite this general trend, a small number of scores show a large divergence between connectivity values for similarly integrated lines (especially i ≈ 1.5).
Fig. 6.12

Moore House, intelligibility graph

6.4.4 Kramer House, Norco, California, USA (1953)

The Kramer House , which was designed for a physician and his family, is approached from the south side and the entry is through a landscaped garden alongside the carport (Fig. 6.13). This southern side of the house was designed to accommodate a guest room that was deliberately isolated from the main bedroom. This guest room was intended to serve the doctor as a second bedroom so that he would not disturb his wife when he made night calls to patients (Lamprecht 2000). The house retains some of the pinwheel properties of the earlier plans, including a centrally located communal space, even though it is not as extensively developed as the plans of the earlier designs (Fig. 6.14). With only fifteen defined spaces, the Kramer House is one of Neutra’s smaller designs of the era.
Fig. 6.13

Kramer House, perspective view

Fig. 6.14

Kramer House, annotated plan

The Kramer House axial map—much like that of the Moore House —reveals the longest and most integrated lines intersect at shallow angles in a dedicated circulation space (Fig. 6.15). Also, like the Moore House, one of these lines passes through the main living spaces at the extreme edge of the area, rather than centrally, as it does in the Kaufmann and Tremaine houses. In the Kramer House the main functional spaces are sited to the side of the circulation core and provide a degree of flexibility by accommodating a secondary circulation system. Indeed, the house features separate public and private flexible zones with the main functional areas located on two of the circulation loops. The master bedroom contains the remaining loop, which is linked by only a single connection to the rest of the house.
Fig. 6.15

Kramer House, axial map

The axial map for the Kramer House features only twenty-two lines (Table 6.7). Significantly, the top third most integrated lines all converge into the same hallway zone. Furthermore, the most integrated line forms the only connection between the master bedroom circulation loop and the remainder of the house (i = 3.75). The second most integrated line (i = 3.46) runs the length of the central hallway from the living room through to the garage. Conversely, the least integrated paths include those to the dressing room or walk-in-wardrobe (i = 1.00) and the courtyard (i = 0.703). In an interesting reflection of the brief, the same line integrates both the master bedroom (B1) and the guest bedroom (B3), separating the two, but keeping them both attached to the house in a similar social structure. This could well reflect the client’s stated desire to use the guest bedroom when he was ‘on-call’. From the formal entrance only a single direction change is required to reach the master or guest bedrooms. While there is geographic or dimensional distance between the entry and the bedrooms, several long, clear vistas and paths connect these spaces.
Table 6.7

Kramer House, results

 

MD

RA

RRA

i

Line length

Connectivity

Lowest three i values

4.0476

0.3047

1.4222

0.7031

10898.2800

1

3.1428

0.2142

1.0000

1.0000

3219.5430

2

3.0952

0.2095

0.9777

1.0227

1753.6630

2

Median three i values

2.2380

0.1238

0.5777

1.7307

13187.5000

4

2.2380

0.1238

0.5777

1.7307

15063.5200

5

2.1904

0.1190

0.5555

1.8000

9005.9490

3

Highest three i values

1.9523

0.0952

0.4444

2.2500

8140.2310

7

1.6190

0.0619

0.2888

3.4615

25993.3100

11

1.5714

0.0571

0.2666

3.7500

19800.3100

12

Min

1.5714

0.0571

0.2666

0.7031

1753.6630

1

Mean

2.3939

0.1393

0.6505

1.7765

10121.1200

4.3636

Max

4.0475

0.3047

1.4222

3.7500

25993.3100

12

Steps distance (E → B1)

1

Total lines in map

22

An R2 value of 0.8965 confirms that the Kramer House is highly intelligible (Fig. 6.16). Thus, it is possible to develop a spatial awareness of the majority of this plan by passing along a relatively small number of paths. This supports Neutra’s general principle of strong spatial orientation; a quality associated with the efficient discovery of the inhabitable spaces in a plan.
Fig. 6.16

Kramer House, intelligibility graph

6.4.5 Oxley House, La Jolla, California, USA (1958)

Neutra described the Oxley House as a ‘modest home’ (1971: 38), with only fifteen defined spaces including several exterior areas. Designed for the family of a physicist, the house is located on a site overlooking the Pacific Ocean in La Jolla (Fig. 6.17). Originally intended to be a pinwheel plan, one wing, containing the pool and a terrace, was never constructed, leaving the house midway between the radiating and linear strategies Neutra typically employed. The entrance is by way of a stepping, angled path that connects to a covered terrace (Fig. 6.18). Once within the entry hall, the visitor is within a space with an angled axis made up of a sequence of doorways which separate five pairs of spaces: bedroom 1 and dining room, dining room and living room, living room and hall, hall and bedroom 2, and bedroom 2 and courtyard. This intricately planned vista (which deliberately angles across the otherwise orthogonal space of the house) is identified by the axial map as the most integrated, and longest, vista and path in the design; a finding which perfectly aligns Neutra’s intentions with the results of this analysis.
Fig. 6.17

Oxley House, perspective view

Fig. 6.18

Oxley House, annotated plan

The axial map displays the dominance of two almost parallel lines, one through the centre of the configuration (as previously noted) and the other connecting an external circulation zone (Fig. 6.19). A single perpendicular line crosses these two, along with three angled ones. The intersection zone between the lines, which generally corresponds to the living, dining and terrace areas, also corresponds to the set of the most integrated lines in the house. Functional spaces supporting the living areas lie on two of the three loops, offering flexibility of use for daily activities. The third loop includes the guest room and garage. Each of these circulation loops requires a degree of external passage in much the same fashion as the Kaufmann Desert House . The master bedroom, living room, hall and guest room all lie on the most integrated line (i = 5.38) (Table 6.8). This line is at the core of the circulation structure of the house, forming part of all of the non-trivial loops. From the formal entrance only a single direction change is required to traverse the most integrated line to either of the bedrooms. Conversely, the least integrated lines are found in the master bedroom en-suite (i = 1.04).
Fig. 6.19

Oxley House, axial map

Table 6.8

Oxley House, results

 

MD

RA

RRA

i

Line length

Connectivity

Lowest three i values

2.8235

0.2279

0.9607

1.0408

8934.4736

1

2.8235

0.2279

0.9607

1.0408

3260.4805

2

2.4705

0.1838

0.7748

1.2906

4693.9688

2

Median three i values

1.8823

0.1102

0.4648

2.1510

10852.956

4

1.8235

0.1029

0.4339

2.3046

10653.231

5

1.8235

0.1029

0.4339

2.3046

19355.965

6

Highest three i values

1.5882

0.0735

0.3099

3.2265

17029.711

9

1.5294

0.0661

0.2789

3.5850

12304.671

9

1.3529

0.0441

0.1859

5.3775

22429.455

11

Min

1.3529

0.04411

0.1859

1.0440

2360.481

1

Mean

1.9738

0.1217

0.5131

2.3020

12181.9776

5

Max

2.8235

0.2279

0.9607

5.3775

23281.6500

11

Steps distance (E → B1)

1

Total lines in map

18

An R2 value of 0.8858 confirms that the Oxley House is also highly intelligible (Fig. 6.20). However the graph is deceptive in that the high integration of the central line would suggest that a greater number of connections are required to more closely match the trend-line. Nevertheless, a review of the design data and map reveals that the majority of internal spaces are directly observable when traversing this line, a quality which in itself would make the house intelligible.
Fig. 6.20

Oxley House, intelligibility graph

6.5 Conclusion

The first hypothesis outlined in this chapter holds that long sight lines will dominate the network of vistas and paths in each of Neutra’s plans. The five axial maps demonstrate that this is true, with a small number of long sightlines shaping and controlling the planning and social structures in all of the designs. These lines typically bisect the heart of the plan and incorporate, or are adjacent to, most of the major functional spaces. Furthermore, when traversing the paths defined by these lines, peripheral vision would allow a person to gain a large amount of information about any remaining parts of the plan. Mathematically, Neutra’s most integrated lines often cross the geometric cores of their plans, and if his architecture is dominated by long vistas as the theory suggests, then a relatively tight integration range along with a low mean result might be evidence for this. Of the five houses, the larger ones typically possess the smallest range of integration values, with the smallest house (Oxley) having the widest range. Only the Moore House , which is separated into two pavilions, is the exception to this rule. However, overall the mean integration values are all low compared to the maximum score in each house. An examination of the full mathematical results confirms that in each design there are a small number of highly integrated paths or vistas, and a large number of lines with significantly lesser values. Overall, the combined quantitative and qualitative data tends to support the first hypothesis.

The second hypothesis holds that Neutra’s plans will possess a high level of cognitive clarity and therefore, that intelligibility results in at least four of the five houses will be greater than R2 = 0.75. The Kramer (R2 = 0.90), Oxley (R2 = 0.89) and Kaufmann (R2 = 0.88) designs all posses very high intelligibility scores but the Moore (R2 = 0.6783) and Tremaine (R2 = 0.55) houses fall below the R2 = 0.75 benchmark. Therefore three of these houses possess a strong sense of orientation and spatial clarity but the other two do not. Interestingly, past research has demonstrated that Neutra’s 1929 Lovell ‘ Health’ House has a particularly labyrinthine plan with a commensurately low level of intelligibility, R2 = 0.2124 (Ostwald and Dawes 2012). Thus, as logic dictates, just because a design has a similar aesthetic expression, it does not mean that its planning is also consistent. Because the second hypothesis was framed as an expectation that at least four of the houses would demonstrate this property, it is disproved. Nevertheless, there is some support for this property in the data.

The third hypothesis has two parts. First, it suggests that even the most private spaces in the plan will be topologically close to the exterior. The shortest path from the most public space to the most private averages only 1.2 direction changes across the five designs. Four of the five houses have a step distance of one! This confirms that the houses are topologically shallow, a finding that is supported by the mean depth results, 1.9 < MD < 3.7. The second part of this hypothesis holds that the structure of the plan will require exterior connections. All five houses have similar planning and circulation strategies, with Neutra consistently including kitchen, living, dining and garage areas on major circulation loops. With the exception of the Kramer House , all of these loops also include external spaces. Moreover, in the case of the Kaufmann and Oxley houses, all loops require external circulation and for the former, three of the four wings are only accessible by way of external paths. On balance, both parts of this hypothesis are confirmed by these results.

If we view the complete set of results, they provide a level of evidence that Neutra created spaces where vision leads to movement and where close proximity to the environment is emphasised and even celebrated. However, despite his stated intentions, experience does not necessarily lead to understanding in Neutra’s planning, with only three of the five cases supporting this contention. Overall, despite not emphatically supporting all three hypotheses, this chapter has uncovered several important traits of Neutra’s architecture that have rarely been taken seriously in the past, but may begin to explain why designers and historians continue to be fascinated by it to the present day.

Copyright information

© Springer International Publishing AG, part of Springer Nature 2018

Authors and Affiliations

  1. 1.School of Architecture and Built EnvironmentThe University of NewcastleCallaghanAustralia
  2. 2.School of Architecture and Built EnvironmentThe University of NewcastleCallaghanAustralia

Personalised recommendations