These considerations arose from a recent conversation I had with a colleague architect about the reception of the concept of space in architecture before the modern epoch. I hope they can help avoid possible misinterpretations where, in a previous article — Concepts of Space in Vitruvius —, I discussed the possibility of interpreting space (spatium) in a three-dimensional sense in relation to certain passages in Vitruvius’s De Architectura. Yet this question is more subtle than it may appear, and it deserves further elucidation, especially for those interested in the relation between space and architecture, and more broadly, in the interpretation of space across different epochs.
To suggest that people before the modern age may have grasped or even developed a proto-modern, three-dimensional notion of space does not mean that such a concept was necessarily applied to architecture as ‘architectural space’. The fully-fledged three-dimensional understanding of reality as space was itself a late development in the gradual evolution of the term. The specific application of space as an architectural concept came even later.
For this reason, I will often stress the use of dimensional qualifiers (one-, two-, or three-dimensional) alongside the word space. When we moderns speak of space, we immediately presuppose a three-dimensional extent. In past epochs, however, that was not always the case. As I noted in the article on Vitruvius, spatium in Latin often referred to one-dimensional, two-dimensional, or temporal extents. I also highlighted the danger of anachronistically applying our modern presupposition of three-dimensional space to earlier thought in my article Concepts of Place, Space, Matter, and the Nature of Physical Existence, based on Ivor Leclerc’s The Nature of Physical Existence. Leclerc, former president of the Metaphysical Society of America and a noted interpreter of Alfred North Whitehead, also wrote a brief essay entitled ‘Concepts of Space,’ where he explicitly observed that ‘until the seventeenth century, the word “space” had the general meaning of “extent”, and in English, back to the fourteenth century, was used in two main specific senses, one in regard to time, an extent or lapse or interval of time, and the other in respect of linear distance, an extent or interval between two or more points, and consequently also a superficial extent or area.’[1] This agrees perfectly with the study I presented on Vitruvius’s varying uses of spatium.
Alongside these historical and linguistic considerations, it is important to distinguish the three-dimensional sense of space from the intuitive grasp of reality as multi-dimensional. Human beings have always experienced reality as a layered, embodied event or happening. However, this event, which is physical reality, is not ‘space,’ but place: we perceive places and the objects or bodies that constitute them, not ‘space’ or ‘spaces.’ On this website I have also explored the possibility of understanding objects or bodies themselves as places — the places of processes that constitute their being (see Being as Place: Introduction to Metaphysics – Part One, and Part Two). In this perspective, architecture — understood as the built artifact — is best conceived not primarily as a question of space, but as a question of place. Space is a reductive term, abstracting only the dimensional aspect (its physical extension) from the fuller reality of place (which has many more colors and nuances —qualities—than extension alone).
For me, the intuitive, three-dimensional, phenomenological interpretation of reality is always rooted in places and the objects within them — what we call the physical environment. This environment, in its natural constitution, has nothing to do with ‘space’. To ascribe the multidimensionality of reality to space is anything but immediate or intuitive, for we do not perceive space. Space does not exist as an independent entity or framework. As the psychologist James J. Gibson once remarked: ‘space has nothing to do with perception’.[2] Space is a pure abstraction, a concept forged over centuries of metaphysical, physical, and mathematical reflection before it could be clearly recognized and enriched with further connotations — phenomenological, existential, artistic, geographical, social, architectural, political, economic, and more. Space, in short, is an invention: a concocted notion, ephemeral and intangible, or, again in Gibson’s words, ‘a myth, a ghost, a fiction for geometers’ (see also James J. Gibson on the Concept of Space).[3]
That is space. If space is a conceptual invention, the crucial questions then arise: when was it invented, and by whom? The articles on this website attempt to shed light on possible answers, but those answers can only be appreciated within a very long narrative that began with the early Greeks and unfolded across many epochs and conceptual frameworks: to apeiron, to kenon, topos, chōra, anaphes phusis, pneuma, aether, spatium, vacuum, locus, absolute space… until more modern and contemporary notions of place and space emerged, such as relative spacetime or the field-concept.
space has nothing to do with perception. We perceive places and the objects, or bodies in them; we do not perceive space, which we can only imagine, as an abstract, immersive frame of reference
Regarding the concept of space in architecture, and specifically the understanding of ‘architectural space’ as an instrumental, three-dimensional construct available to architects for designing, this conception only began to emerge in the late nineteenth century and became a widespread domain of architectural knowledge only in the second half of the twentieth century (see the Appendix in the article On the Ambiguous Language of Space). Returning, then, to the interpretation of space in Vitruvius: the Roman architect might have developed a conception of space — and, specifically, of architectural space — in a one- or two-dimensional sense (as linear distance, area, region, surface, etc., see Images 04, 05, 06 in Concepts of Space in Vitruvius). Yet it is doubtful that he could have envisioned a clear-cut three-dimensional conception of space as a background containing objects, walls, floors, roofs, people, and so forth; and, above all, it is highly improbable, if not impossible, that he could have conceived such a ‘background space’ as an instrumental tool for architecture — that is, as three-dimensional architectural space. Let us look at this question more closely.
In antiquity, the Greek atomists developed the notion of to kenon (the void), which bore some analogies with the modern concept of space, though it was not identical to it. The Greeks, in fact, had no explicit concept of space, nor a single term for it. Instead, to kenon, topos, and chōra were the main terms by which they described what we might now call the dimensional or environing character of reality — the ‘the cosmic scene of all events’, the continuum binding everything together in both physical/material and temporal respects.[4] Depending on whether this continuum was conceived as empty, occupied, or traversed by bodies, they spoke of kenon, topos, or chōra. Other Greek terms also described this all-pervading substratum, such as anaphes phusis, pneuma, or aether, each carrying specific material or spiritual connotations. None of these terms, however, can be directly translated as ‘space’. Moreover — as Heidegger observed — while we may retrospectively emphasize the dimensional character (extensio) of this cosmic scene, the early Greeks did not experience it as mere extensiveness.[5]
The atomistic void (to kenon) and its Epicurean development as an ‘intangible substance’ (anaphes phusis)[6] were later elaborated by the Roman poet Lucretius, a contemporary of Vitruvius, in De Rerum Natura.[7] In translating Greek concepts into Latin — to kenon into vacuum, spatium, or [locus] inane — Lucretius introduced new nuances. Yet these were exceptional cases in antiquity: those who privileged void or space were relatively rare, compared to those who emphasized the more concrete notion of place. Indeed, the dominant referential concept for the cosmic scene of events was not space or void, but place (topos, sometimes chōra for the Greeks; locus for the Romans). Aristotelian cosmology and Ptolemaic astronomy ensured the dominance of place over rival atomistic or Stoic accounts. Importantly, no sharp distinction was drawn between void (vacuum), space (spatium), and place (locus) until between the sixteenth and eighteenth centuries (a point I examined in Concepts of Place, Space, Matter, and the Nature of Physical Existence).
Thus, while it is not impossible that Vitruvius was aware of Lucretius’ elaborations and glimpsed a proto-modern conception of space (spatium) as a generic, three-dimensional expanse, it is highly improbable that he translated this into a fully instrumental concept for architecture. The transfer of such an idea would not only have required transposing a philosophical notion (from natural philosophy) into the architectural domain, but also scaling it from the microscopic domain of atomism to the mesoscale of human affairs, i.e., architectural scale.
Here lies the central point. We cannot rule out the possibility that Vitruvius, as a Roman intellectual, following Lucretius, may have entertained the idea of spatium as a three-dimensional expanse — as a kind of protomodern conception of space. This possibility underlies certain passages of De Architectura (notably 5.9.5 and 6.6.3), where I left open the interpretation of spatium in a three-dimensional sense, even in contexts describing architectural environments (especially 5.9.5)—see the article Concepts of Space in Vitruvius. Similarly, in other passages (catalogued in the article under ‘figurative gap or cavity’), I suggested a possible three-dimensional reading.
Yet it is crucial to distinguish between understanding space as a generic three-dimensional expanse, gap, or cavity, and conceiving space as an instrumental design tool. The leap from perceiving three-dimensionality to employing ‘3D space’ as a medium of architectural design requires another level of abstraction — as shown by developments after Galileo, Descartes, Newton, and, later, by aesthetic and philosophical debates from the late nineteenth century onward. For architecture specifically, this leap involves a kind of epistemological and phenomenological epiphany. It is not something to be taken for granted. I have illustrated this with my own early encounter with ‘space’ (which eventually crystallized into architectural models I have termed Archi-textures; see the related article), and, by analogy, with Paul Klee’s famous epiphany with colour on his way to Kairouan in 1914, when he wrote in his diary: ‘Colour possesses me. I don’t have to pursue it. It will possess me always, I know it. That is the meaning of this happy hour: Colour and I are one. I am a painter.’[8]
Image 01: Paul Klee, Before the Gates of Kairouan, 1914, watercolour on paper
One question concerns what we have before our eyes; another concerns the attribution of meaning to what is before our eyes (or, in the case of space, to what is not even visible). In this regard—what is given to perception, as distinct from the meaning attributed to it—I am reminded of a passage in Investigations, by the American theoretical biologist and complex systems researcher Stuart Kauffman. He recounts a curious episode from the Spanish conquest of the Americas: the native inhabitants gave no response to the arrival of the first Spanish ships in the Caribbean, because ‘the ships were not seen — there was no concept for them.’[9]
These observations lead to an important conclusion, one I have also discussed in connection with the articles on The International Style (Mind, Space, Architecture: On The International Style and On the Ambiguous Language of Space): to read the architectures of past epochs—Greek, Roman, Renaissance, Baroque, Neoclassical—in a spatial, three-dimensional sense may indeed be useful for modern architects and students. Yet it cannot be the primary way of accessing their meaning and value. For the interpretation of premodern architectures, material, volumetric (in the sense of mass), geometrical, and aesthetic considerations (disposition, proportion, symmetry, etc.) must precede spatial or metrical ones. A three-dimensional spatial conception could not have guided the hands and minds of architects in those epochs; they lacked such a tool. To think otherwise would be an anachronism.
Overall, it seems to me that three distinct spatial objects are implicated here:
Type 1. Three-dimensional space understood as a neutral void or background for events, bodies, and things.
Type 2. Three-dimensional space as architectural space—an initially neutral expanse (Type 1) that takes on the qualities or atmosphere of the built object (walls, roofs, windows, balconies, fences, etc.). This is essentially a passive transformation of space into architectural presence, whether interior or exterior.
Type 3. Architectural 3D space-as-a-tool—the most abstract and complex type. This space has topological, geometrical, and symbolic properties that can be extended into aesthetic, functional, and sociocultural dimensions. It is the space of the project (a thoughtful creation of the human mind), a model through which architects can anticipate and shape the perception and cognition of a built environment.
The difference among these types reflects different levels of engagement. Types 1 and 2 are more immediately accessible, as seemingly external structures ‘observable’ from the outside. What we actually perceive, however, is always a place, filled with bodies or objects, and their relations; ‘space’ as such does not exist as an external object of perception. Type 3, by contrast, belongs to creation: it presupposes an awareness not only of the environment as a three-dimensional fact, but also of the mechanisms of perception and cognition themselves, allowing the architect to influence them in advance through the spatial/volumetric structure of a project. This possibility is strictly modern. Earlier architects could only manipulate material elements—walls, windows, roofs, symmetries—without the support of an abstract spatial tool. For this reason, architectural space-as-a-tool (Type 3) was unknown to Vitruvius and to all premodern architects, as attested by the articles focused on The International Style and related discussions (again, I redirect readers to Mind, Space, Architecture: On The International Style and On the Ambiguous Language of Space).
Now, of course, architects have conceptions of space that go beyond three-dimensionality, reaching into relativistic spacetime or the field concept. But these, especially the field concept, should be reinterpreted within the broader category of place, which is the core of the redefinition I am pursuing at RSaP—Rethinking Space and Place.
To conclude: in De Architectura, space had prevailing one-dimensional, two-dimensional, and temporal meanings. Vitruvius’s architectural understanding of space was, at most, two-dimensional (e.g., 6.2.5—see Concepts of Space in Vitruvius). On rare occasions, he may have used spatium in a generic three-dimensional sense (in the Lucretian, Epicurean sense), describing a figurative expanse, gap, or cavity (Types 1 and 2, above). This confirms three points:
- Vitruvius understood space, in architectural terms, at most, as a two-dimensional concept.
- There is a crucial difference between conceiving space as three-dimensional and using it as a tool for design. The latter requires another level of abstraction.
- Vitruvius may have glimpsed a protomodern sense of three-dimensional spatium as a container of bodies, but not the notion of architectural space-as-a-tool.
This defines the state of the art concerning the traditional notion of architectural space—the concept established from the late nineteenth century onward, excluding more recent work on non-Euclidean or relativistic spaces. A promising direction for further research would be to investigate whether this state of the art is definitive, or merely the result of insufficient interdisciplinary study of premodern authors. What is needed is not more historical accounts that risk anachronism, but rather close analysis of the spatial language used by architects such as, for instance, Brunelleschi, Alberti, Palladio, Perrault, or Ruskin, in comparison with the prevailing concepts of space and place in their respective epochs, both physically and metaphysically or philosophically. Such research would require collaboration between architects, attuned to the nuances of spatial terminology, and philosophers or historians with expertise in spatial questions.
Notes
[1] Ivor Leclerc, ‘Concepts of Space’, in Probability, Time, and Space in Eighteenth-Century Literature, ed. Paula R. Backscheider (New York: AMS Press, Inc., 1979), 209-216.
[2] This is the extended quotation ‘the concept of space has nothing to do with perception. Geometrical space is a pure abstraction… The visual third dimension is a misapplication of Descartes’s notion of three axes for a coordinate system. The doctrine that we could not perceive the world around us unless we already had the concept of space is nonsense. It is quite the other way around: We could not conceive of empty space unless we could see the ground under our feet and the sky above. Space is a myth, a ghost, a fiction for geometers.’, in James J. Gibson, The Ecological Approach to Visual Perception (New York: Psychology Press Classic Editions, 2015), xv.
See also James J. Gibson on the Concept of Space.
[3] Ibid., xv.
[4] I have taken the expression ‘the cosmic scene of all events’ by Samuel Sambursky, slightly modifying the original formulation — ‘the cosmic scene of material events’ — which was an expression referred to the Stoics’ concept of ‘pneuma, an all-pervading substratum’; in Samuel Sambursky, Physics of the Stoics (Princeton: Princeton University Press, 2016), 1.
[5] Martin Heidegger, Introduction to Metaphysics, trans. by Gregory Fried, and Richard Polt (New Haven: Yale University Press, 2000), 69. See also note [111] in the article Being as Place: Introduction to Metaphysics – Part One.
[6] Because of this concept, ‘intangible substance’, Epicurus is said to be the first to have isolated the concept of space. It this regard, in Casey’s book The Fate of Place, we find a quotation, elaborated on a statement from Long and Sedley, according to whom Epicurus was ‘the first ancient thinker to isolate space in the broadest sense’. In a passage from Sextus Empiricus and reported in Casey’s The Fate of Place, Epicurus refers to an ‘intangible substance’ (anaphes phusis), as he calls it, which is named ‘void’ (kenon), ‘place’ (topos), or ‘room’ (chōra) according to whether that substance is empty of all body, is occupied by a body, or bodies roam through it. In: Edward S.Casey, The Fate of Place: A Philosophical History (Berkeley: University of California Press, 1997), 83.
[7] An interesting contribution concerning this question and, more in general, concerning the reception of the concept of space in the Roman period is the paper by Carlos Lévy, — ‘Roman Philosophy under Construction: the Concept of Spatium from Lucretius to Cicero’; here, the discernment between the different meanings of space-spatium is especially focused on the ‘chronological’ versus ‘local’ sense, without putting special emphasis on the distinctions concerning the different dimensionalities intrinsic to the ‘local aspect’, e.g., ‘it seems that in pre-Lucretian Latin there was a rather clear separation between temporal aspects, or more precisely durative aspects, mainly expressed by “spatium”, and local aspects, for which “locus” was generally used’; Carlos Lévy, ‘Roman Philosophy under Construction: the Concept of Spatium from Lucretius to Cicero’, in Space in Hellenistic Philosophy: Critical Studies in Ancient Physics, eds.G.Ranocchia, C. Helmig, and C. Horn, (Berlin, München, Boston: De Gruyter, 2014), 129.
[8] See also note [11] in the article Concepts of Space in Vitruvius.
[9] Stuart Kauffman, Investigations (New York: Oxford University Press, 2000), 81.
Works Cited
Casey, Edward S. The Fate of Place: A Philosophical History. Berkeley: University of California Press, 1997.
Gibson, James J. The Ecological Approach to Visual Perception. New York: Psychology Press Classic Editions, 2015.
Heidegger, Martin. Introduction to Metaphysics. Translated by Gregory Fried, and Richard Polt. New Haven: Yale University Press, 2000.
Kauffman, Stuart. Investigations. New York: Oxford University Press, 2000.
Leclerc, Ivor. ‘Concepts of Space’. In Probability, Time, and Space in Eighteenth-Century Literature, edited by Paula R. Backscheider, 209-216. New York: AMS Press, Inc., 1979.
Lévy, Carlos. ‘Roman Philosophy under Construction: the Concept of Spatium from Lucretius to Cicero’. In Space in Hellenistic Philosophy: Critical Studies in Ancient Physics, edited by: Graziano Ranocchia, Christoph Helmig and Christoph Horn. Berlin, München, Boston: De Gruyter, 2014.
Sambursky, Samuel. Physics of the Stoics. Princeton: Princeton University Press, 2016.
Vitruvius. ‘The Ten Books on Architecture’, translated by Morris Hicky Morgan. London: Humphrey Milford Oxford University Press, 1914.
Image Credits
Featured Image: Paul Klee, View of Kairouan, 1914, 73, watercolour and pencil on paper laid down on cardboard, 8.4 x 21.1 cm (Franz Marc Museum, Kochel am See, permanent private loan).
Image 01: Paul Klee, Before the Gates of Kairouan, 1914, 216, watercolour on paper on cardboard, 20,7 x 31,5 cm (Zentrum Paul Klee, Bern)
For further information about the images, see also the digital resource Mapping Klee offered by the Zentrum Paul Klee.
