Introduction

Paul Valéry (1871–1945) was a major figure in French cultural life in the first half of the twentieth century. Although widely read, and translated, during his lifetime, his classical, conservative and anti-modernist tastes have not been to his favour, and his reputation went into eclipse following his death. Yet there is much in Valéry’s thinking that was radical, and anticipated aspects of late 20th and early 21st-century currents of thought (Gifford and Stimpson 1998: 2). This is particularly evident in the Cahiers, the series of 259 notebooks compiled by Valéry over the course of his lifetime, but not published until later.¹ They reveal a more intriguing and diverse thinker than he had previously been given credit for, and since the French publication of the Cahiers, first in facsimile in the late 1950s and then in edited form in the early 1970s, his work has undergone a major re-assessment, though this has been largely confined to literary circles, with less impact in the visual arts (Figure 1).

Figure 1

Paul Valéry, Self Portrait. 1935. One of a number self-portraits in the Cahiers. Paul Valéry, Cahier no. 165: 11r. BnF, Paris.

Valéry’s primary vocation was literary, but his interests ranged widely, and included architecture, which was, as he says in the Cahiers, his first passion:

My first love was architecture, both of ships and of buildings on land. The possibility of constructing space according to solid forms; the way of imagining the structures, the sustained heights, the interweaving forces playing against each other; the art of casting over a passageway or a stone covering, the art of leading from one face of the structure to another, the intoxication of tectonics, of perceptible relationships.. it possessed me completely. — The picturesque, in the first instance — then I cured myself of that to devote myself to the organic.

I still don’t know how to express exactly what I loved about constructions. I think I found in them somewhat vaguely the idea of noble deeds, — of machines — of superhuman movements which have been made into reality. […]

I loved to think about buildings and found them mentally stimulating. It made me feel good to imagine them. (1918, C/N 1: 116)

Architecture occupies a significant, though elusive, place in his thought. Throughout his writings there are many references, direct and indirect, to architecture, although he wrote only one published text explicitly about architecture, the dialogue Eupalinos or the Architect of 1921. Almost all previous attention to Valéry’s architectural thinking has been concentrated upon this one piece, regarded by the German critic Hans Blumenberg (1964) as one of the most important theoretical art and architectural texts of the 20th century. Eupalinos was widely read by architects in the early to mid-20th century, especially in the French-speaking world — by Auguste Perret, Le Corbusier, Fernand Pouillon, amongst others (Britton 2001; Foucart 1995; Maak 2011; Bonillo 2001). Valéry wrote Eupalinos as a commission for a sumptuous volume on architecture in a projected but never realised serial encyclopaedia published in a small print run in 1921; it was reprinted in a smaller, much cheaper, unillustrated format in 1923 (Karaïskakis and Chapon 1976: 48–55). It is not an easy text, and English language readers are not well served by the rhetorical style of the only English translation. A Socratic dialogue set in Hades between two shades, Socrates and Phaedrus, the conversation moves through different sequences elucidating the theme of thinking versus constructing and the possible differences and analogies between them. Reference will be made to Eupalinos here because in it Valéry amplified ideas in the Cahiers, but it is not the purpose of this article to elucidate Eupalinos, or its more recent commentaries (Cacciari 1980; Bekaert 2011). In previous studies of Valéry’s treatment of architecture, the Cahiers, if not absent, have been treated as marginal (Pelmont 1949; Bouret 1970; Foucart 1995; Jallat 1982; Maak 2011; Signorile 1993; Signorile 2009). For this article, the situation is reversed, and the Cahiers are taken as the primary source, to which other texts are secondary, and in what follows we explore the Cahiers for what they might reveal about Valéry’s conception of architecture. The first part introduces the Cahiers and their method for the benefit of an architectural audience that may not be familiar with them, and then looks at the place of architecture within the wider scheme of the Cahiers. The second extracts a number of ‘architectural’ themes from the Cahiers, and examines each of them in turn. No attempt has been made to situate Valéry’s thinking more broadly within the intellectual or political circumstances of his time, and it is only the architectural context that is addressed here, and that only partially. While any discussion of Valéry’s work is almost bound to fail in preserving the full coherence of his thought, its scale and ambition should not be lost sight of.

The Cahiers

The Cahiers were a series of notes written by Valéry in the very early hours of almost every morning from the early 1890s up to his death in 1945. They are like no other work in modern times; they have been compared to Leonardo da Vinci’s Notebooks, which Valéry had studied and contemplated writing about. Discontinuous and often idiosyncratically punctuated, they range in length from the very short to longer entries of up to two or three pages (Figure 2). These are not ‘notes’ in the sense of the kinds of memos, lists or sprawls that one may usually associate with notebooks or a daily journal. As Paul Gifford describes them, the Cahiers should not be mistaken for a ‘private record of personal hopes, fears and aspirations, no chronicle of deeds and days […] He [Valéry] is not out to please, to fascinate or dominate anyone’. Rather they are ‘a kind of laboratory of thought in which is being elaborated the most central of human sciences, elucidating and containing all others’ (C/N 1: 23). As Gifford writes elsewhere, there is also something very consistent about them:

The fragmentary character of the notebooks is one of the greatest challenges to acquired habits of reading. Yet it is also key to the endless fascination of this unique document. The Valéryan fragment is a natural pulse or unit of mental energy: enough to stimulate, enrich and direct the mind without exhausting it. Rewarding in its own right, it avoids closure in the guise of rigid development — something Valéry most deeply mistrusted in systems and ideologies. The mind, in his view, is authentic only in its rebounding, ever-renewed, always unfinished act; as achieved totality of understanding, it overreaches and falsifies itself. At the same time, as we come to know the thought of the Cahiers more intimately, the patterns of interconnection between fragments reveal an underlying consistency that is both massively coherent and immensely supple in operation; we begin to see and to savour the unitary mind informing the cosmos of fragments. (Gifford 1998: 38–39)

Figure 2

A characteristic page from the Cahiers containing notes on a variety of topics, including two on works of art: ‘A work is part of an act — no existence in itself’. ‘The sole interest of art (the work) is to draw out of us what we didn’t know we contained — and did not contain’. 1935. Paul Valéry Cahier no. 160: 55r. BnF, Paris.

While the Cahiers were never published in his lifetime, Valéry regarded them as his principal work. And yet he avoided calling them a ‘work’, preferring instead to call them ‘counter-works, counter finites’, because, as he explained, ‘As far as “thought” is concerned, works are falsifications, since they eliminate the provisional and the non-repeatable, the instantaneous and the mingling of purity and impurity, disorder and order’ (1937–38, C/N 1: 48).

Alongside the written notes, Valéry included sketches and diagrams, and the combination of the two is an important part of the rhythmic and interrupted thinking of the Cahiers, as is the placement of these elements on the page (Ryan 2007). For Valéry, drawing was another way of thinking, a counterpart to writing, a means of looking, puzzling out how things work and deducing a structure, function or relationship. A full investigation of the images in the Cahiers vis-á-vis architecture would be a major undertaking, requiring a method very different to the textual analysis employed here. Although Paul Ryan’s 2007 study of Valéry’s drawing practice provides a precedent, he did not consider the architectural dimension. The pages reproduced in this article have been chosen to show the range and variety of the graphic material with architectural potential in the Cahiers, and the captions provide some commentary on them.

During his lifetime Valéry made three separate attempts (in 1898, 1908 and 1921) at organising the Cahiers thematically, but none was ever fully carried out. The third scheme would have arranged the entries in 31 categories — ‘Art and Aesthetics’, ‘Gladiator’, ‘Poetry’, ‘Ego’, ‘Mathematics’, ‘BIO’, ‘History and Politics’, ‘Eros’, ‘Psychology’, ‘Education’, ‘Science’, ‘Memory’, etc. — and these are the categories used in the edited editions. Architecture was not one of these categories; references to architecture are distributed throughout. To take the entries on architecture out of their context is to lose sight of their place within the whole and risks violating the fluidity and open-ended nature of the Cahiers — one has to remain constantly aware of the greater project. While others have mined the Cahiers for Valéry’s ideas about particular practices or themes (e.g. music, shells), this is an enterprise that has to be undertaken with caution (Stimpson 1984, 1988; Pietra 1991). It has not been attempted systematically up to now for architecture.

Method

The project of the Cahiers — which has been explained and substantiated very fully elsewhere, so here is only summarised — was an all-encompassing, universal quest to reformulate thought through a concentrated effort and exercise of how the mind can think about itself without falling back on pre-existing systems, philosophies, or theories of mental activity (Gifford 1998). To do this Valéry relied upon his mind, the only means available to him: ‘Other people write books, I am making my mind’ (1902–3, C/N 1: 67). He was well aware of the difficulties and the contradictions. The project that became the Cahiers is partly prefigured in the essay ‘Introduction to the Method of Leonardo da Vinci’ (1894), where Valéry sketched out some of the limitations ‘about the situation of and the means available to a mind that sets out to imagine a mind’ (CW 8: 4). This same essay is also the other main published work in which Valéry formulated ideas about architecture, and for that reason also, we shall make reference to it here.

The paradox of the Cahiers is that while Valéry was adamant about rejecting all methods, there is nevertheless method to them. When he first began writing notes in the early 1890s, the goal was to create an all-comprehending ‘System’ of knowledge; a unitary model for thinking from which all possible variables (all thoughts and connections of thoughts) and its products could be derived — for which he drew on different disciplines such as thermodynamics and mathematics. This ‘system’ could be focused upon any image, idea, sensation, etc., to identify the different connections that might exist between mental states:

Though it’s a considerable risk, I am attempting what Faraday attempted and achieved in physics, Riemann in mathematics — Pasteur in biology — and others in music. I am attempting to give the theory of consciousness a method which is sufficiently rigorous to restrict the number of ghosts it contains and to establish closer connections between the practical branches it has always possessed, irrespective of successive theories. (1900, C/N 5:41)

This was not an assertion of the mind over the body, nor of intellect over emotion. Valéry considered the function of the mind from both conscious and unconscious responses, including dreams and sleep and the power of the imagination and of memory (the past made present). He was only too aware of the fragmentary nature of his thought processes — ‘I have a unitary mind in a thousand pieces’ (1900–1, C/N 1: 62); the form of the Cahiers respects that while also trying to overcome it. Around 1900 Valéry began progressively to abandon the ‘system’, and the ambition of a definitive end-product (C/N 5: 11–15).

Despite abandoning his earlier ambitions of a ‘system’, there are consistencies across the notes. He returns over and over again to the same questions, and reflections upon the mind, rethinking them in a process that has been compared to Penelope’s weaving (C/N 1: 21). Thus, towards the end of his life he reflected:

I write these notes a little as one practices scales — and they’ve been repeated on the same notes for fifty years —, rather as one goes for a walk at such and such a time — each day. And I write them not to make some work or system out of them, but as if I was going to live indefinitely, fulfilling a stationary duty – just as a spider spins its web without a future or past, just as a mollusc would continue to produce its spiral shell – not seeing why or how it might cease to secrete it, step by step. (1940, C/N 1: 50)

Likewise, throughout the fifty years of writing, there is a wariness of his own mind, and a striking degree of self-reflexiveness: ‘the greatest power of thought should be directed against the power of its thoughts — all of which are suspect’ (1939, C/N 4: 291).

Valéry’s project of constructing his mind might be seen as comparable to Freud’s investigation of the unconscious and to the contemporary development of psychoanalysis. Although Valéry was dismissive of Freud, as ‘unscientific’, and was in general resistant to looking for meanings (‘I “see” phenomenally — i.e. reduced to appearance, — without depth, — without meaning […] others […] fly from the thing to the meaning without taking sufficient account of what is, and thus reaching too quickly what is not, the effect of what is’ (1926, C/N 1:139)), there was, as has been pointed out, more in common between their thinking than Valéry was prepared to admit (Bowie 1998, C/N 3: 9–10, 20). There are points of similarity, such as a respect for the insignificant, the trivial — ‘For the pure intellect, nothing is trivial’ (1914, C/N 1: 269).

One of the inherent contradictions of the Cahiers was his dependence upon language to articulate thoughts and sensations that were invariably pre-linguistic (and this was one of his complaints about psychoanalysis — ‘You relate what you felt before the time you had the words to speak’ (1941, C/N 3: 228)). Valéry, despite his vocation as a writer, was constantly suspicious of language as a medium, for its tendency to constrain the mind. (Nonetheless, it should be said, for all his misgivings about language, his own use of it is economical and exact; he chooses words with great care, and he is never rhetorical.) Furthermore, the nature of mental activity is in its constant movement and flux, which language can never replicate, tending always to freeze a thought; what he was after, as he put it, was how to recognise that ‘The operations of the mind can best serve our purposes of analysis while they are still moving, unresolved, still at the mercy of a moment’ (CW 8: 9). In so far as Valéry aspired to a ‘method’, while at the same time avoiding any method, he was inspired by the example of science, where emphasis is on questions or problems over solutions or answers: ‘A real science is not a system of answers. On the contrary, it’s a system of problems which remain forever open. The fundamental axioms of a science are partial determinations of the problems’ (1901–2, C/N 4: 229). Over time, he became increasingly resistant to arriving at answers: ‘There’s always something in me that makes me prefer not to have a solution’ (1935, C/N 1: 180).

Others have scrutinised the ‘method’ of the Cahiers more closely, but for the moment it is enough to stress their combination of an inexhaustible attention to particular problems, combined with Valéry’s instinct ‘to dread the definitive’ (1905, C/N 1: 76); a refusal to accept unquestioningly pre-conceived schemes of thought (even, ultimately, those of science); ‘a preference for the reserved, the unstated’ (1938, C/N 1: 195); a resistance to a spurious clarity — ‘learn to take due account of obscurity’ (1923, C/N 1: 280); a predilection to ‘to be precise about generally vague things and vague about generally precise things’ (1931, C/N 1: 159); and his continuous self-questioning of his own way of looking at everything he turns his attention to (Laurenti 1979; C/N 5: 11–15). Valéry’s advice to ‘go at your business obliquely’, which appears in the section ‘Gladiator’ (1923, C/N 1: 280), concerned with how to train the mind (as one might a horse), is a recurrent theme of the Cahiers.

‘Architecture (how close to the ways of my mind!)’

References to ‘architecture’ are spread across the Cahiers (Figures 3, 4, 5). However, though there are many sketches of buildings and of interiors, remarks about existing built works of architecture are relatively rare and mostly relate either to Gothic architecture, about which, inspired by his early reading of Viollet-le-Duc, he was enthusiastic, or to modern architecture, which he deplored. Gothic he liked for the way it gave expression to hidden forces within it, and for its correspondences with plant forms. Modernism he disparaged for various reasons, but the following is characteristic:

Architectural style —

No two works of art are the same. Yet, what is achieved in so-called modern architecture is quite the opposite.

What once used to distinguish one cathedral from another or one house from another was at the very least, the ornamentation — —

Production-line architecture is like printing — Print-run of x copies. (1944, C/N 2: 89)

Figure 3

House in the countryside near Cuers. 1933. With Fig. 4, one of the rare cases instances where Valéry draws an actual building for the purpose of analysing its architecture. Paul Valéry, Cahier no. 152: 38v. BnF, Paris.

Figure 4

Krakow, Jesuit church. 1936. Paul Valéry, Cahier no. 172: 9r. BnF, Paris.

Figure 5

Rome. ‘I don’t like the ruins’. 1937. (The Victor Emmanuel Monument is visible through the window). Valéry quite often drew the room in which he had been sleeping. Paul Valéry, Cahier no. 175: 2r. BnF, Paris.

If these were Valéry’s private thoughts, his public self was more open to contemporary developments. He was gratified that architects took notice of what he wrote, and is known to have conversed and corresponded with Auguste Perret and Le Corbusier (1941, C/N 2: 522; Britton 2001; Le Corbusier 1987).

Previous studies of Valéry’s treatment of architecture have generally focused upon the early formation of his interest, and his responses to actual buildings or styles of architecture, but in the Cahiers these are the least interesting of his remarks (Bouret 1970; Foucart 1995; Maak 2011). Far more insightful are the occasional suggestive notes about architecture in general, such as in these two examples, both from 1894:

As I see it, a building such as we know it is only a nucleus: architecture for me also means forms that have never been used — and attention on the other hand to known styles, not following the usual processes, but taken in whatever way takes your fancy. (1894, Ed. Int., 1: 387)

The second is from the published essay ‘The Introduction to the Method of Leonardo da Vinci’:

Architecture is commonly misunderstood. Our notion of it varies from stage setting to that of an investment in housing. I suggest we refer to the idea of the City in order to appreciate its universality, and that we should come to know its complex charm by recalling the multiplicity of its aspects. For a building to be motionless is the exception; our pleasure comes from moving about it so as to make the building move in turn, while we enjoy all the combinations of its parts, as they vary. (CW 8: 51)

There is a contemporaneous diagram in the Journal de Bord, the first of the numbered series of the Cahiers, that apparently represents something of the same thought (Figure 6).

Figure 6

‘Life’. 1894. Suggestive of the plurality of the possible experiences of architecture within the city. Paul Valéry, Journal de Bord: Cahier no. 1: 32v. BnF, Paris.

Another, slightly later entry in the Cahiers offers a far from orthodox view of the arts, in terms that could have been those of William Morris: ‘One can imagine an “era” when the specialised, deliberate arts would be abolished and replaced by the art of ordinary activities. In short, by the art of living’ (1917, C/N 2: 45).

Each one of these remarks is astonishing for its date, and belies Valéry as a conservative. The first would seem to imply that he was open to forms of architecture that were unimagined, and that existing monuments were simply a germ of what might be possible. The second suggests a dynamic view of architecture the likes of which were not to be found in architectural circles of the time, nor indeed until the radical architecture movements of the 1960s. The third could be read as anticipating the post-war writing of Henri Lefebvre. The discrepancy between the conventionality of his tastes and the radicality of his ideas occurs in many aspects of his thought, as well as in relation to the visual arts (Adorno 2019; Cooper 1960). It is one of the many surprises Valéry can spring on us.

Why did architecture appeal to Valéry? Here the Cahiers offer several insights. In the first place, part of its attraction lay in its not being like language, of whose inadequacies Valéry was, as we have seen, painfully aware. Not only was architecture inherently non-signifying (1933, C/N 2: 72), a view expressed in Eupalinos, but also, as he wrote,

I’ve always been tempted by activities which denied one the use of language — — such as music and architecture. My sensibility only enjoys constructions and modulations — which literature only offers in a forced and miserly way, for language requires a constant change of object, without repetition. (1943, C/N 1: 244)

The strongest appeal of architecture, though, seems to have been that it offered a way to describe, or to picture, the operations of the mind — an idea that is repeated many times, in different forms. In his 1931 annotations to ‘The Introduction to the Method of Leonardo da Vinci’, he writes that ‘thinking is only an effort to pass from disorder to order’ (CW 8: 28), the same definition of architecture and construction that he gives elsewhere. This analogy between architecture and the mind recurs implicitly throughout the Cahiers, and is fundamental to them: as he acknowledged, ‘Architecture (how close to the ways of my mind!)’ (1923, C/N 2: 50). The analogy is a pliable one, and allowed for extension into a variety of aspects of thought (Figure 7).

Figure 7

‘The mind, envisaged as enclosure— (in my “mind”). The relation of interiority and exteriority, between two unknowns’. 1935. The mind represented as a house, with various flows (‘light’, ‘liquids’, ‘gasses’, ‘gravity’) connecting it to the outside world. Paul Valéry, Cahier no. 160: 76r. BnF, Paris.

Of the recurrent references to architecture in the Cahiers perhaps the most persistent relates to Valéry’s ongoing pre-occupation with what distinguishes the creations of the natural world from those of humans. Everything that the mind produces is a human artefact, yet the mind itself is a natural organ, so the boundary between the two is less distinct than we might like to think. Architecture represented for Valéry one of the most complex and complete manifestations of human creativity, and it thus plays an important part in the pursuit of this question of the differences between natural and human creation. Valéry’s preference for medieval Gothic was in part because it contained a structural system embedded in the ornament, not necessarily visible on the surface, but which one nevertheless knows to be there — and in that respect like the mind (Robinson-Valéry 2010: 72). But more than this, as he explains,

Architecture known as Gothic, is the only sort in which material, structure and ornament are (very often) in complete harmony. Even with the Greeks, something comes between them. The Gothic is ‘generative’ inasmuch as an artistic action can imitate the ‘generative’.

It is vegetative. (1942, C/N 2: 87)

‘Generative’ here recalls other notes in the Cahiers:

Beautiful architecture is somewhat plant-like […] The law of growth must be felt. Just like the law governing the layout of openings. — A window must not be a hole, like something drilled into a plank of wood, but the culmination of internal laws, like the mucous membrane and contours of natural orifices. Around the points of change modifications must occur — For every building a rule governing the whole set of modifications must be established.

Analogy consists of imitating not the forms themselves, but transposed laws. (1917, C/N 2: 44–45)

Medieval Gothic appealed because it represented a freedom of the mind, as of the body, that Valéry aspired to in his thought.

The bird and its extraordinary activity completely intoxicate me. Ultra sensitive, ultra swift, superbly darting swallows.

I am madly envious of this mobility.

[…]

A land-based animal is similar to an architectural construction which experiences only its load-bearing function, the accumulated weight of elements composing it; whereas animals like the fish or the bird, which require of themselves ceaseless action, are like a gothic construction. (1914, C/N 1: 268–69)

But however close Gothic might be to the natural order, it was not of it, and architecture’s distinctness from everything in nature remains a cause of fascination to him: by what means can the human mind have conceived this building, any building, and have brought it into existence?

Construction

If ‘architecture’ is an elastic concept in the Cahiers, so too, to an even greater degree, is ‘construction’, as will already be evident from what has been said so far. Sometimes synonymous with ‘architecture’, at other times not, Valéry’s understanding of ‘construction’ ranges from building, to literary composition, to painting, to the organisation of men for some purpose, and is for him one of mankind’s greatest gifts — ‘the riches […] the broad intellectual domains that are illuminated by the conscious act of constructing’ (CW 8: 41). Later, in Eupalinos, he is even more categorical: ‘of all acts the most complete is that of constructing’ (CW 4: 145). His introduction to his 1931 play Amphion underlines the centrality of construction to him:

My young self passionately imagined the act of construction […] and even the idea of construction, which is the transition from chaos to order, the use of the changeable to arrive at necessity, took hold of me as the most beautiful type of activity that a person can imagine. A finished building gives us, at a single glance, the show of intentions, inventions, insights and forces that imply its existence; it brings to light humanity’s combined work of desire, knowledge and ability. Uniquely among the arts, and an indivisible moment of vision, architecture charges our souls with the sense of human capacities as a whole. (Bekaert 2011: 48)

Valéry conceives the act of construction in various different ways. In ‘The Introduction to the Method of Leonardo da Vinci’, it involves the substitution of one thing by another:

Constructing takes place between a project or a particular vision and the materials that one has chosen. For one order of things, which is initial, we substitute another order, whatever may be the objects rearranged: stones, colours, words, concepts, men etc. (CW 8: 41)

If the first sense of ‘construction’ refers to the transformation of chaos into order, or of one type of order into another, a second sense arises in Valéry’s pre-occupation with the distinction between the human and the natural products. While both are ‘constructions’,

Nature’s constructions of living things doesn’t make use of materials in the same way as man. It can’t prepare homogenous metallic masses; whereas man can’t construct an object on a human scale from molecular or atomic measurements.

In order to establish an object of sufficient size, nature will set up a network of colloids into which it will introduce mineral elements, by way of solution and precipitation.

It can’t make use of anything that is not or was not soluble. Hence the importance of salts.

What’s more, it can’t distance itself from its work.

(1914, C/N 4: 135)

A third sense of ‘construction’ is closely connected to looking. Sight is particularly important in Valéry’s scheme. When it comes to architecture,

The pleasure offered by architecture is visual in the first place — but should then extend to the tactile, the mobile, the rhythmic, the surprising — to the idea of the material of the building.

1st instant — sight. (1930, C/N 2: 65)

Looking, however, is not a purely passive act but an activity complementary to, and in some sense the inverse of, that involved in the making of the work: ‘it is by a process of abstraction that works of art are constructed […] Inversely, it is by a sort of induction, by the production of mental images, that works of art are appreciated’ (CW 8: 46). In other words, looking is itself a constructive act, by which a work’s coming into being is played out in reverse. The act of seeing is ultimately of the same nature as the act of constructing: ‘He who looks constructs […] Architecture as an art is the art of recognising how to construct a system of figures conforming to the nature of the gaze that excites the very need that it satisfies’ (unpublished MS, quoted in Ryan 2007: 263).

A fourth sense of ‘construction’ relates to language. In Eupalinos, Phaedrus asks, ‘And so language is a constructor?’, to which Socrates responds ‘most certainly the word can build, just as it can create, as it can corrupt’ (CW 4: 106), and goes on to elaborate a parallel between modes of language and modes of architecture. Yet if language is constructive, there is nevertheless a difference between knowledge of the kind known to philosophers through language and that revealed by construction. Later in the dialogue, Phaedrus says ‘I now understand how you could hesitate between constructing and knowing’ (CW 4: 124). Compared to philosophy, of which Valéry was generally dismissive, regarding it as too speculative and as having failed to come up with worthwhile, verifiable insights, ‘construction’ offered access to a different, non-linguistic, kind of knowledge.

Finally, there is a sense in which ‘construction’ relates to the self. In Eupalinos, Phaedrus quotes Eupalinos: ‘“By dint of constructing,’ he put it with a smile, ‘I truly believe that I have constructed myself”’. To which Socrates responds by asking, ‘“To construct oneself, to know oneself — are these two distinct acts or not?”’ (CW 4: 81). Such a process of constructing the self underlies the whole of the Cahiers, which, for all their apparent formlessness and fragmentariness, were the procedure through which Valéry set about just this task. ‘I don’t construct a ‘System’ — My system — is me’ (1942, C/N 1: 235), he writes at one point, underlining the total identification of what is constructed with the self. Construction is an act, through which one draws out what was previously hidden — so, in relation to the work of art, he writes, ‘The sole interest of art (the work) is to draw out of us what we didn’t know we contained — and did not contain’ (1935, C/N 2: 142). But this action reflects the features of the interior self, so ‘Every construction (not determined by exterior local circumstances) is made according to the sensibility — i.e. by contrasts, likenesses, symmetries — diverse forms of generalized repetition’ (1936, C/N 2: 143).

The act of ‘drawing out’ results in the realisation of a form. As Valéry wrote in a letter to a friend shortly after completing Eupalinos, ‘I should have tried to show that pure thought and the search for truth in itself can only ever aspire to the discovery or the construction of some form’ (CW 4: xii).

Form encompasses not just architecture, but any man-made or natural-made thing. In pursuing construction, one constructs oneself, or finds out about the process of thinking itself — but the translation is not a straightforward one from one ‘function’ to another, for Valéry never makes such simplistic analogies, and the project of the Cahiers is too diverse to be contained within one single image.

Works

Where the ‘work’ of a work of art resides has been a long-standing, perplexing question in aesthetic philosophy, and nowhere more so than in relation to architecture (Macarthur 2024: 73–106). Although Valéry did not engage with this legacy, he offers a very particular perspective on the issue. One does not get far in the Cahiers without coming up against what Valéry has to say about ‘works’. Valéry’s thoughts about ‘works’ take many different turns, and are central to his aesthetics. Although most of Valéry’s discussion of ‘works’ is irrespective of any particular medium, and can be taken to apply equally well to literary, pictorial, architectural or musical productions, nevertheless his trade as a writer inclined him to give rather more attention to literary works, though even here much of what he says relates equally well to other art forms. We should bear in mind that where English has one word, ‘work’, French has several — ‘oeuvre’, ‘ouvrage’, ‘travail’, ‘labeur’, all used by Valéry — making possible opportunities for nuance and discrimination that are not so readily available in English. Where appropriate, we have indicated the word used by Valéry.

As already seen in relation to his project for the Cahiers, Valéry was dismissive of ‘works’, and was inclined towards forms that did not qualify as works. His objection to works is repeated many times, in different variants, but always with the same intent: ‘For me, a work of art is not a completed being, sufficient unto itself — it’s the cast-off skin of an animal, a cobweb, a deserted conch or shell, a cocoon. It’s the creature and its work that exercises me’ (1913, C/N 2: 105). Or, more succinctly, ‘The works of man seem excrement to me — residues of acts. I love them only in order to imagine the acts of their formation’ (1916, C/N 1: 108) (Figure 8).

Figure 8

‘The work as made is a sort of excrement’. 1937. A page apparently exploring the characteristics of a ‘work’, whether natural, or human. Paul Valéry, Cahier no. 181: 88r. BnF, Paris.

But at the same time as being generally derisory of works, he does nevertheless concede them a purpose, which may be to serve either the author or the audience, or both. ‘The purpose of a work [oeuvre] is to astonish its creator [son ouvrier]’ (1912, C/N 2: 103). And in relation to the audience, ‘The objective of a work [ouvrage] — an honest work — is simple and clear: provoke thought’ (1912, C/N 2: 463). It is emphatically not, however, the purpose of works to deliver a meaning. Valéry always draws back from meaning. As he writes about himself, ‘My faculty — readily seeing things as devoid of meaning’ (1903, C/N 1: 69). In relation to art, the most ‘poetic’, for him, is that where there is ‘the impossibility of reducing the thoughts it provokes to any finite, exhaustive, dismissive meaning‘ (1937, C/N 2: 82). ‘A work that is produced has no true meaning, and the author is no more capable of legitimately and reliably revealing it than anyone else’ (1926, C/N 2: 302).

Part of the quality that Valéry looks for in works is that they should not be exhausted in their usefulness, that they should retain a reserve against their exposure to audiences: ‘The poem does not die from services rendered; it is expressly made to rise again out of its ashes and to become endlessly what it has just been’ (1939, Œ 1: 1331). There is a sense that the work has within it a surplus, a surplus that is, or should be, inexhaustible. And this idea connects to the way that Valéry sees the making of works: ‘A work of art [une oeuvre] is the residue of labour [travail]’ (1934, C/N 2: 76). The idea that works contain more within them than is apparent, that they are the outcome of an excess of energy, is one that we shall encounter again in relation to ornament. Works result from a superfluity beyond the needs of pure utility:

It’s the inexact adaptation of the animal to its environment, the ‘slack’ left over — already apparent in the waste of seed, in that death indispensable to life (and primarily in the $\frac{\mathrm{99..}}{\mathrm{10..}}$ of seeds (spermatozoa)) which shows the discrepancy clearly — the fissure in the life-machine, through which the Mind will pass — that breakthrough towards the hazardous adventure of consciousness and the creation of works. (1942, C/N 2: 152–53)

It is in the ‘residue’, the traces of productive excess that Valéry looks for in works, that he finds the sense of the poetic.

While Valéry avoids the pursuit of meaning, audiences can be brought into an engagement with works in other ways. Works are what activate readers, viewers, audiences, and in Valéry’s aesthetics, the engagement of readers is no less important, indeed in some respects more important, than the labour of authors: ‘It is never the author who creates a “chef-d’oeuvre”. A chef-d’oeuvre depends on its readers, on the quality of the reader […] He alone can create a chef-d’oeuvre’ (1916, C/N 2: 267). Valéry’s ideas about readers in many senses anticipate those of Roland Barthes’s 1968 essay ‘The Death of the Author’ (1977). Occasionally he calls them ‘lecteurs — exécutants’, ‘readers — performers’, underlining their role in the making and constant remaking of the work (1922, C VIII: 875).

Few people imagine one can produce a work specially written — not with a view to giving the reader something, but with a view to receiving. To offer the reader the chance of a pleasure — active work [travail] — instead of proposing a passive enjoyment. A text fabricated expressly to receive a meaning — and that not only one meaning, but as many meanings as the operation of a mind upon a text can elicit from it. (1933, C/N 2: 138)

There’s a lot more in the reflections occasioned by a text than in the actual thought that produced it. (1937, C/N 2: 328)

Some of these ideas about authors and audiences are summed up in more general terms in his ‘Opening Lecture of the Course in Poetics’ at the Collège de France in 1937: ‘the mutual independence of the producer and the consumer, their ignorance of each other’s thoughts and needs, is almost essential to the effect of a work’ (CW 13: 98).

Despite all his reservations about works, those of music and of architecture turn out to be partially exempt from the misgivings he has about other arts. It is worth considering why they escape the objections he has against other sorts of works. Part of the appeal of both architecture and music to Valéry is that in them questions of meaning are less intrusive than in other art forms.

Music and Architecture make us think of something quite other than themselves; they are in the midst of this world like the monuments of another world […] they invoke the constructions of the mind, and its freedom, which is in search of this order and reconstitutes it in a thousand ways; they therefore neglect the particular appearances with which the world and the mind are ordinarily occupied: plants, beasts and people. (CW 4: 98)

What distinguishes music and architecture from other arts is their capacity not to tell stories, but rather, as he puts it, ‘to engender in us not a fable, but that hidden power that makes all fables’ (CW 4: 97). Painting and poetry lack this virtue (Figure 9).

Figure 9

‘The painter searches for painting, etc. Each art searches for what it alone can do, what can only be done by it, and nothing else than that. Poetry is the most embarrassed in this regard because of the ambiguity of language’. 1937. As is often the case, the relationship between the drawing and the adjacent note is uncertain. Paul Valéry, Cahier no. 181: 20v. BnF, Paris.

Architecture and music are the purest manifestations of the process of ‘construction’, of the casting of thought into a form through precise relations. More than any other art or science, Valéry sees them as exceptions to his general strictures about works, and especially architecture lends itself to the kind of retrospective construction, the seeing of the many steps that the maker has gone through, that Valéry so appreciates. Some of this was articulated in the ‘Introduction to the Method of Leonardo’:

It is through the monument or, one might rather say, among such imaginary scaffoldings as might be conceived to harmonize its conditions one with another […] that we are best able to reconstitute the clear intelligence of a Leonardo. Such a mind can play at imagining the future sensations of the man who will make a circuit of the edifice, draw near, appear at a window, and by picturing what the man will see; or by following the weight of the roof as it is carried down walls and buttresses to the foundations; or by feeling the balanced stress of the beams and the vibration of the wind that will torment them; or by foreseeing the forms of light playing freely over tiles and cornices, then diffused, encaged in rooms where the sun touches the floors. It will test and judge the pressure of the lintel on its supports, the expediency of the arch, the difficulties of the vaulting, the cascades of steps gushing from their landings, and all the power of invention that terminates in a durable mass, embellished, defended, and made liquid with windows, built for our lives, to contain our words, and out of it our smoke will rise. (CW 8: 50–51)

Long though it is, this passage expresses how in architecture, to a degree not evident in the other arts, the work more nearly exposes the thought processes of the maker, and it is this that causes Valéry to concede to architectural works a privileged status within his thinking.

Valéry’s ideas about the ‘work’ are not unitary. Running through them is the sense that a ‘work’ is not the end of work. A work exists despite itself, rather than because of itself, and is potentially more than its physical existence might lead us to suppose. This is not to say that Valéry was in favour of the ‘unfinished’. Contrary to what is sometimes supposed, Valéry was not a champion of the ‘non-finito’ — on the contrary, he found the cult of the sketch, its elevation to a ‘work’, absurd, ‘taking for an end what can only be a means’ (CW 12: 20). On the other hand, for Valéry no true work was ever complete, either for its author or for its audience/reader: ‘What is perfect imposes incompletion. Perfect is that which you cannot have done with and which eludes the mind while at the same time enticing it’ (1937, C/N 2: 547n69). As regards ‘works’, this is perhaps the message from Valéry that is most relevant to the practice, and criticism, of architecture.

Making

If Valéry was sceptical of ‘works’, he very much valued ‘making’. There are many notes in the Cahiers reiterating the significance of ‘making’.

The act of making matters more to me than the object.

The work [l’ouvrage] is in its making, which is the crucial goal, as I see it […]

One ends up by placing the act of the worker as more important than the work itself. (1927–28, C/N 2: 127)

And he speculates whether the act of making of a work might not take the place of or become the work itself: ‘The real work of art is in its composition, or in its fabrication. Turn this stage into a work of its own??’ (1920–1, C VII: 700). However, even in relation to literature, Valéry does not show any particular curiosity in the Cahiers towards the specific techniques of how works are made, in the details of execution, or in craft as such. This comes as a surprise given his enthusiasm for ‘making’, but it would seem that his concern was more to do with what takes place in the interval between a mind’s imagining and the realisation of a result rather than in the particular skills employed in any given work (Figure 10).

Figure 10

‘Consummation is much more difficult to represent than production’. 1941. A diagram exploring visually a relationship between two ideas. Paul Valéry, Cahier no. 214: 43r. BnF, Paris.

As with ‘works’, a large part of the interest in ‘making’ is concerned with the contrast between nature and man. Art ‘starts with the need to make’, leading to an act in which pleasure lies, but this is not the way nature makes (1930, C/N 2: 64). Not only is nature more complex in its composition than man-made things, and this is discussed at length in Eupalinos and in his 1936 essay L’homme et la coquille, but there is also in man-made things a separation of thinking from doing:

The tree does not construct its branches and leaves; nor the cock his beak and feathers. But the tree, and all its parts, and the cock and all his, are constructed by the principles themselves, which do not exist apart from the constructing. That which makes, and that which is made, are indivisible. (CW 4: 126)

The indivisibility for Valéry is a fundamental difference between the things created by nature and those made by mankind, which, although a product of nature, cannot make anything as nature does. However intuitive and iterative human production may be, it invariably requires prior thought and a reliance upon processes separate from ourselves. Human making begins in the mind. Nature is a non-thinker when it ‘makes’ itself. A too easy correlation between the two, endowing the natural with intentionality, is misguided:

Note. It’s the generality and the crudeness of the notion of Making which has engendered cosmology. If, on the other hand, we oblige ourselves to specify the actions of making and their means — we make it impossible for ourselves to think (in all vagueness) that an animal or a world are fabricated like a vase or a clock. NB. The word create is often used for ill-defined fabrications. . without any apparent model. (1930, C/N 5: 273)

Discussing a seashell (objects Valéry was fascinated by) in L’homme et la coquille, he writes that while we might be able to conceive of a shell and replicate its form, we cannot mimic its unthinking growth:

The idea of making is the first and most human of ideas. ‘To explain’ is never anything more than to describe a way of making: it is merely to remake in thought. The why and the how, which are only ways of expressing the implications of this idea, inject themselves into every statement, demanding satisfaction at all costs. Metaphysics and science are merely an unlimited development of this demand. They may even lead us to pretend not to know what we know, when what we know refuses to be reduced to a clear knowledge of how to make something.… This is what we mean by going back to the beginnings of knowledge.

Here then I will introduce the artifice of a doubt: considering this shell, in whose shape I think I can discern a certain ‘construction’ and as it were the work of some hand not acting ‘at random’, I ask myself: Who made it? (CW 13: 12-13)

In architecture, where biological analogies and metaphors have been commonplace, Valéry’s insistence on an absolute distinction between man-made and natural artefacts would still seem productive and worth holding on to (Wigzell 2020) (Figure 11).

Figure 11

One page among many of drawings of shells in the Cahiers. 1936. Paul Valéry, Cahier no. 168: 71v. BnF, Paris.

Ornament

Formative in Valéry’s teenage reading was his discovery in the Montpellier municipal library of Owen Jones’s The Grammar of Ornament, first published in 1856. Jones’s book showed, in sumptuous graphic form, that ornament, in all its variety and proliferation, was based upon a few general principles, set out in 37 propositions. The realisation that beneath the apparent chaos of ornamental schemes from all over the world an order could be found appealed to Valéry, because, as Judith Robinson-Valéry (1998) has argued, he saw there a correspondence with mathematics. Surprisingly for so unornamental writer, one who deplored eloquence and rhetoric, he retained a life-long fascination with visual ornament (Jallat 1982: 254–82; Ryan 2019: 61–79).

His earliest discussion of ornament occurs in the 1894 essay ‘Introduction to the Method of Leonardo da Vinci’, where it has to be understood in the context of a broader concern, with how the mind is capable of finding order amidst all the variety and irregularity of the world.

The world is irregularly strewn with regular arrangements. Crystals are of this nature; so are flowers and leaves, many striped or spotted ornaments on the fur, wings, or shells of animals, the patterns made by the wind on matter or sand, etc.

[…]

The knowledge of regular combinations is divided among the different sciences; or, where none can be established, it comes under the calculation of probabilities. For our purpose we need only the observation made on introducing this topic: that regular combinations, whether of time or space, are irregularly distributed in the field of investigation. Mentally they seem to be opposed to a vast number of formless things. (CW 8: 28–29)

In his later gloss on this passage, Valéry adds, ‘If everything were irregular, or if everything were regular, there would be no thinking — for thinking is only an effort to pass from disorder to order’ (CW 8: 28).

Ornament represents a very particular instance of the mind’s capacity to perceive order out of the variety and irregularity of things: ‘Ornament, that answer to emptiness, that compensation of the possible, in a sense completes and is also the annulment of a liberty’ (CW 8: 43-4). Ornament takes known objects, but ‘their meaning and ordinary use are ignored so that only their order and mutual reactions may subsist. On this order depends the effect’ (CW 8: 45). In this respect, ornament is the outcome of a mental act, a projection applied to irregularities and variations present in nature.

In the Cahiers, Valéry returns to this question. Sometimes he simply enthuses about a particular ornament or style of ornament (he particularly liked Arabic ornament) (1936, C/N 2: 79); at other times, he is more analytical. He makes a distinction between ‘décor’ and ‘ornament’:

When the architectural effect is above all one of shock — and no longer allows the passage from the detail to the whole and from the whole to the detail – it is merely decor — decor can be exalting — create a certain mood — inspire — but does not complete — choir — but the effect does not last. (1930, C/N 2: 544)

Décor, it would seem, belongs in the same class as ‘luxury’, which he saw as no more than ‘waste material’, and of which he disapproved, as akin to drunkenness or over-stimulation (1921, C/N 1: 124; 1944, C/N 2: 152). Ornament, on the other hand, is always connected to the active mind.

Ornament consists of creating materially, of materializing, what the eye would create by itself, if it had the power to project its ‘desires’. It must be given what it would give itself, reduced to its own modes of pleasure and suffering, of renewal and conservation, of increased intensity/quantity/of its free energy. (1920–21, C/N 2: 48)

Valéry here points to what is the most original aspect of his theory of ornament, namely that it is the result of a projection of mental ordering on to a given material, ‘a kind of “spontaneous “ creation, local to the sensory organs […] hearing, sight, etc.’ (1925–26, C/N 2: 124). This notion of the psychological origin of ornament closely anticipates that elaborated later by Ernst Gombrich in The Sense of Order. Although Valéry had planned a history of ornament in 1891–92, he never developed his concept of the psychological basis of ornament, though it was clearly something that continued to occupy him (Jaillat 1982: 278–80). He recounts an evening spent with Henri Bergson, during which ‘I expound my theory of ornament — the formation by the sensibility of something to fill the void — according to local laws — general laws (contrasts — symmetries)’ (1929, C/N 1: 152). Ornament comes about both through a desire to fill a void — ‘Ornament — filling emptiness — empty time — empty “space” which nature abhors’ (1935, C/N 2: 78) — and as a result of ‘Surplus of energy. Unused freedom — The mark of a power greater than the principal action requires and absorbs — Cf. children’s games’ (1905–6, C/N 2: 163). Seen in these terms, an ornament might be no more than a gesture,

The closing smile of the acrobat or of someone who has just finished some difficult piece of work, as if to offer it free of charge and to say it’s of no consequence, quite easy, as light as a feather […] This smile, the demonstration of independence, and this freedom as ornament of a highly important thing, or of a trivial thing which is nevertheless dangerous or awkward. The ornament arises here as a sign, and as a sign of superiority, of superabundance […] I am what is necessary, plus something else. (1923, C/N 1: 281)

The dual nature of ornament, both as coming from within and as a response to a void, Valéry sums up in a late entry:

Ornament is a filler (when it exists alone) which responds to a sensation of emptiness (Add. Emptiness as a positive sensation) — and I dare compare this to the rhythmic singings, unconscious hummings, automatic gestures — which are produced in the individual when he is unoccupied or preoccupied, like the expenditure of usable, excess energy. (1944, C/N 2: 154)

Intriguingly, Valéry sometimes regrets this filling up of emptiness — ‘the annulment of a liberty’, he wrote in his 1930 annotations to ‘The Introduction to the Mind of Leonardo’ —nevertheless his regard for voids was not a quality he was prepared to extend to contemporary architecture, whose lack of ornament he generally deplored (1944, C/N 2: 89).

One of Valéry’s English translators, Robert Pickering, writes that ‘The status of ornament is a central theme in Valéry’s thought on aesthetics and art’ (C/N 1: 619n31). This may well be so, but what exactly that place was remains elusive. Undeveloped as the theme was, his ideas are suggestive, and in certain respects anticipate later writing about the theory of ornament.

Modulation

Possibly the best known, and most often repeated, architectural maxim in Valéry’s Eupalinos is ‘there are no details in execution’ (CW 4: 71). Paraphrased in one form or another by Le Corbusier, Auguste Perret and many others, it was variously interpreted to justify an absence of ornament; standardisation; or the quality of seamlessness in a good work of architecture. In fact, what Valéry meant by the phrase was not the same as his architectural readers supposed. The example Socrates gives in the dialogue is of the surgeon who, skilled as he may be, uses a dirty thread or needle, and so kills his patient; and Eupalinos, according to Phaedrus, gave such care to every aspect of his buildings that ‘you would have thought it was his own body he was tending’ (CW 4: 74). For Valéry, the message of the maxim is that nothing is too trivial, everything matters; ‘details’ here are not exactly those in the architectural sense. The conversation in Eupalinos shortly moves on to ‘modulations’ — how things join together, or transition from one state to another, a matter of great importance to Valéry, and one rather closer to what architects have usually meant by ‘details’ (Figure 12).

Figure 12

‘The human face. The dog’s skin slackens and flows over the muscular core — the folds. Story of modulations’. 1923. Valéry was fascinated by modulations of all kinds, in both natural and man-made objects. Paul Valéry, Cahier no. 101: 105r. BnF, Paris.

Modulations, as interpreted by Valéry, are a fundamental feature of life, occurring in the transitions of nature, between sleeping and waking, the lines of a face, the folds of a dog’s skin, and likewise in all art forms — painting, music, poetry, dance, and architecture. They were for him one of the most basic ways through which both natural objects and human artefacts were to be appreciated in common (Jallat 1982: 344–58). In the following passage he introduces the word ‘modenatura’ (which has a French equivalent, modénature, with a specifically architectural meaning, and was the term used by Le Corbusier) to describe the phenomenon, but elsewhere the word he invariably uses is ‘modulation’, which had the advantage of not being tied to any particular practice.

Modenatura.

Passages and modulations — The most refined secret of art — and the mark of exquisite art.

Anyone who doesn’t know this is an ass — However powerful. It links the finger or the hand lightly passing over something and the form of light-and-shade.

Living nature is unbeatable at this. It knows how to terminate a stem, open an orifice, unfurl the end of a canal — extend an external organ, embed an eyeball.

Transitions.

The problem is a deep one — For it involves nothing less than the combination of action with matter (in the relative sense of something that preserves itself) or the opposition and combination of construction with formation (cf. signif[icant] and formal. I still haven’t (after 44 years) got to the bottom of this business). (1937, C/N 2: 1937)

His complaint about dance was that ‘I […] am usually bored when I see it. You don’t have time to read the modulations of the body properly’ (1939, C/N 2: 84).

A recurrent theme was how much better nature manages modulations than do human creations, and the contrast provides one of the main grounds of comparison between the natural and the human. Modulation is a non-exclusive, expansive category present in phenomena of all kinds.

There is nothing better than the joints of plants and those that are readily found in animal organs — to bring home to us the problem of Modulation, which in all the arts is the element of the divine.

To move from one level to another, from one tone to another, from one ‘universe’ to another — this substitution like going from sleep to wakefulness and back — from day to night and back — from being free to being tied —

At the extremities of plants —, the joins, the various differentiations from the same rank or ring or age-level (because the age of a plant is form and mass —). There is a moment of plant-time in which the change in shape becomes more pronounced — a sort of morphogenetic periodic acceleration — growth, shape and functions are linked. (Think of architecture — but don’t think too much of known architecture, of given types and styles. There is also the idea of growth to be taken into account.

Nobody has been able to justify the choices of proportions — etc.)

In lifeless, geological nature, modulation hardly ever occurs except by chance. *At the moment* when a plant, having just sent out a uniform segment, comes to the point of emitting radiant parts, each one of a different form— with colours, functions etc., this is when the irregularity occurs. In this extraordinary act, chemistry, pressure, combinations of forces: tradition — conditions of continuity — Form in equilibrium with age.

The intellect is at a loss when faced with this composition of order.

(1933, C/N 4: 159)

In relation to architecture, modulations play an especially important role because, as a non-signifying art, it is through them that the ‘story’ (although of course Valéry denies there being any such thing in architecture) is communicated. The architect’s whole craft, he emphasises in Eupalinos, relies upon the mastery of how the junctions are handled: ‘imperceptible modulations’, ‘insensible curves’, ‘minute and all-powerful inflections’, ‘deep-wrought combinations of the regular and the irregular’ (CW 4:74). Elsewhere he elaborates on this:

It is the subtle quality of the moldings which will help the builder find the means of transition in hard material, create an almost natural substitution of plane for plane, express the differing functions of the segments and limbs of the building, and display the outward form of the organic unity which will distinguish it from an arbitrary collection of blocks, or a precariously erected pile of packing cases. In this way, he will try to guide the eye by degrees over the effect as a whole, arresting it imperceptibly at one point, to bring it back to another, prompting questions which presently find their answer, thus giving the idea of something produced by life itself. (1938, CW 12: 200–1)

The language of moldings, a long-standing and well-recognised part of the apparatus of the classical tradition in architecture, was rejected by the 20th-century modernists. Contrarily, Le Corbusier argued strongly in Vers une architecture for the significance of modénature, in discussion of the Parthenon, comparing it, in a trope that goes back to 18th-century French architectural theory, to the composition of a face: ‘Contour modulation [modénature] is free of all constraint: it is a total invention that makes a face radiant or withers it’ and ‘is the touchstone of the architect’ (Le Corbusier 2008: 246). Le Corbusier was criticised for using this term, but he stuck to his argument that it was a basic feature of good architecture.

Modénature is the critical point of architecture […] it will determine light and shade, that is to say what the eye sees and what consequently gives us architectural emotion […] Modénature exists in a building without a cornice or moldings: it is its profile, everything concerning its profile,

and he insisted that modénature was not to be confused with traditional moldings (Le Corbusier 1925: 117–18).

Although one of Valéry’s main complaints against modern architecture was that it lacked modulation (1930, C/N 2:65), he did not put his case in terms of a rearguard defence of the classical or Gothic traditions, and there is no historicist bent to his arguments. Valéry’s enthusiasm for modulation accords with Le Corbusier’s privileging of modénature in the ‘axis of harmony’ that ‘ravishes the mind’, but it also prefigures a much more contemporary concern, for it is in relation to modulations that Valéry comes closest to articulating a bio-ethical stance for architecture, a way of judging architecture against the measure of nature, and it is here that his originality lies.

Space

In the 1894 essay ‘Introduction to the Method of Leonardo’, there is a passage on architectural space. Long though it is, it is worth quoting in full because it displays an unusually prescient grasp of this property that has been central to so much 20th-century architectural discourse:

The monument, that creature of stone, exists in space. What we call space is relative to the existence of whatever structures we may choose to conceive. The architectural structure interprets space, and leads to hypotheses on the nature of space, in a quite special manner.

[…]

Space, as soon as we try to represent it in our minds, at once ceases to be a void and is filled with a host of arbitrary constructions; and in all cases it can be replaced by a juxtaposition of forms, which we know can be rendered as required. However complicated we may suppose an edifice to be, it will represent, when multiplied and proportionally diminished in size, the basic unit of a formation of space, and the qualities of the formation will depend on those of the unit. Thus, we can find ourselves surrounded by and moving among a multitude of structures. Let us note, for example, in how many different fashions the space around us is occupied — in other words, is formed, is conceivable — and let us try to grasp the conditions implied in our being able to perceive the variety of things with their particular qualities: a fabric, a mineral, a liquid, a cloud of smoke. We can form a clear idea of any of these textures only by enlarging a single particle and by inserting into it a structure so organized that, by simple multiplication, it will form a substance having the same qualities as the substance under consideration. … With the aid of concepts like these we can move continuously through the apparently quite separate domains of the artist and the scientist, proceeding from the most poetic, even the most fantastic, construction to one that is tangible and ponderable. (CW 8: 52–54)

In this dense passage, we encounter at least three different ideas about the spatiality of architecture, none of which were current in France at the time. We may draw out, in particular, firstly the proposition that architecture interprets space, and can lead to hypotheses about the nature of space. Secondly, space in so far as it is a property of the mind only takes shape through the presence of objects in the perceived world. Thirdly, the objects through which we apprehend space may be of many different kinds — a fabric, a mineral, a liquid, a cloud of smoke — but in every case these substances only acquire a physical reality through their expansion into a space. Valéry recognises that space is both a property of mind but is also what enables substances to become real. Architecture, understood in the broadest sense, is the art through which these various attributes of space are brought into play.

For its date, there was nothing like this in France, no discussion in French architectural culture of architecture as an art of space. Although there was already in Germany a developing discourse around ‘space’ as the defining feature of architecture, it was taking place only within a very limited circle of aesthetic philosophers, and would have been unknown to Valéry. (The first account of recent German aesthetic theory in French was M. Borissavliévitch, Les théories de l’architecture, 1926) (Forty 2000: 256–75; Mallgrave and Ikonomou 1994). How Valéry arrived at these highly original insights remains something of a mystery. While he evidently had a passing knowledge of Kant’s Critique of Pure Reason (1781) and of Kant’s concept of space as an a priori property of mind, he later claimed never to have read Kant closely. He might have come across Nietzsche’s ideas of space as not empty but a ‘force field’, but this is unlikely, since he did not read Nietzsche closely until some years later, after 1898 (C/N 5: 564n2). Valéry did not go on to develop these ideas about architectural space much further, though there is one entry in the Cahiers from 1911 that suggests he had not forgotten them:

Architecture is an ode to space itself. It has to make visible the properties of space and in particular its heterogeneity in relation to man and its homogeneity in relation to the operation of the mind — to virtual movements. (1911, C/N 2: 39)

And in his 1930 commentary on the ‘The Method of Leonardo’, he added the following:

From this point of view his intellectual labors are part of the slow transformation by which the notion of space — at first that of a complete vacuum or an isotropic volume — has little by little developed into the notion of a system inseparable from the matter it contains, and from time. (CW 8: 35)

And, alongside the passage quoted at the start of this section,

This might be the place for a few remarks on space, a word that changes its meaning with one’s angle of vision or manner of speaking. The space of ordinary practice is not quite the same as that of the physicists, which is not quite the same as that of geometricians — since none of these is defined by all the same experiences and operations. (CW 8: 53–54)

With these remarks, we have two further notions of space, firstly that it is inseparable from the matter it contains and from time, and secondly the heterogeneity of space, that it is different things to different people and in different circumstances. Valéry’s notion that space is inseparable from the matter it contains might be supposed to link Valéry to Heidegger, who was developing his ideas about space as not existing independently of the objects within it at around the same time, and later, through his 1952 essay ‘Building Dwelling Thinking’, widely circulated in the architectural world.

However, despite the superficial similarity of their ideas, Valéry’s notions of space turn out to be quite distinct from Heidegger’s. Although Valéry wrote no more about architecture as an art of space, he had a great deal to say more generally about space; along with time, it is one of the two physical concepts that most preoccupied him throughout the Cahiers. While having much to say about it, he is at the same time full of warnings about the instability and unsatisfactoriness of ‘space’ as a term:

Why take as imposed and as well-constructed or even ‘natural’ the notions of time and space which have no more reach or precision than the other words of ordinary language […]

Space is not a clearer notion than time. Sometimes this space is all points and contains nothing but points. No possible movement. Then you place bodies in it, (material points etc.), therefore liaisons. Then the maintenance of these bodies and the variation of their distances, the infinite divisibility and the maintenance of space itself jockey to rational recognition. […]

These notions of space and time are not to be taken seriously —

i.e., don’t reason rigorously using words that do not bear it, and have no vocation for this task. (1916–17, C/N 5: 184)

There are many other entries in the same vein, deploring the reliance of philosophers on the term, and expressing his general mistrust of all abstract nouns (Figure 13).

Figure 13

‘Classic space and time have “popular” origins — that is to say as “solid” bodies and displacement’. 1932. One of Valéry’s many statements about the material basis of space, as a property of physical phenomena. The transition of thought from the diagram to the note beneath is, as often, hard to ascertain. Paul Valéry, Cahier no. 147: 23r. BnF, Paris.

Recurrent throughout is Valéry’s preoccupation with the need to recognise that space can only be understood in relation to time: it is not just the presence, but also the movement of bodies through space that animates space: ‘Space is nothing, neither visible nor definable. When we do speak of it in this way, this is just an allusion to the way in which we distribute sensations and transformations, conservations of some given property of bodies’ (1913, C/N 5: 152). Part of his reason for being so dismissive of Kant was that Kant separates space as a property of mind from the active, lived experience of space that is necessary to its perception. In his emphasis upon the engagement of the observing mind in the embodied experience of space, he distances himself both from Kant and from Heidegger, whose account of space is similarly bloodless. In this respect, Valéry’s ideas of space and spatiality are closer to those articulated by Maurice Merleau-Ponty in Phenomenology of Perception, of 1945: ‘far from my body’s being for me no more than a fragment of space, there would be no space at all for me if I had no body’ (Merleau-Ponty 1962: 146).

Since Valéry’s thinking about space remained within the Cahiers, and was therefore not known about until after his death, one cannot place Valéry within the evolving discourse of space within 20th-century architecture. Nevertheless, he had identified its basic conundrum, the confusion between its being a property of the mind and a physical feature of the world — a confusion which has characterised the entire discussion of ‘space’ in architecture. Valéry recognised that ‘space’ was not so much a property in itself, but rather an idea that answered a need, the need to identify other properties or qualities (1943, C/N 5: 350); this has certainly been true of architecture in the 20th century, where ‘space’ has been used not so much to articulate any definite property but rather as a way of representing what architecture is not, or should not be. Furthermore, Valéry gave architecture a privileged position in so far as he saw it having a unique capacity, not enjoyed by any other practice, to interpret what ‘space’, in all its heterogeneity, might be. Had this been better known, his standing amongst architects might have been even higher than it was.

Coda

A conclusion would be a most un-Valéryian way to end an article about the Cahiers. Valéry was not a thinker with a thesis, as can be said of other writers, like Roland Barthes, Gilles Deleuze or Henri Lefebvre, whose work has sustained architectural thought in recent years. He was instead, as others have remarked, an aphorist, with ‘areas of stubborn interest, of energetic repetition’, and the notes in the Cahiers are provocations rather than systematic arguments readily imported to a discipline like architecture (Raine 2000: 8). Asking what Valéry offers to architecture is less straightforward than in other fields, where he may be said to have anticipated later developments. From our reading of the Cahiers, certain persistencies, certain areas of stubborn interest, emerge. Firstly, the Cahiers were a project about the mind, and architecture serves as an analogy for the mind. Secondly, there is Valéry’s insistence on the non-signifying nature of architecture, and its capacity instead to create ‘another world’ (CW 4, 98). Thirdly, Valéry stresses process over product. Although architecture was a partial exception to Valéry’s general dismissal of completed works in that he took evident pleasure in finished buildings, it was for the sake of them as physical manifestations of a mind’s workings rather than for their sensuous properties. There is a difference here from the more usual understanding of ‘process’ in architecture that focusses on the stages of a design and its translation into a building. Fourthly, they reveal a continuing enquiry into the likenesses and unlikenesses of architecture to organic nature. Fifthly, Valéry validates ornament, and puts forward some original ideas about it. And finally, there is an exceptionally forward-looking enquiry into architecture as an art of space.

Further exploration of the Cahiers (for our reading has been limited largely to the edited versions) might yield other themes or priorities. Valéry is not without his shortcomings and blind-spots. There is for example a blatant disregard for the social, political or economic aspects of architecture, and for architecture’s effects on how we live. Given the importance these issues had within early 20th-century modern architecture, one might expect a public intellectual to have had some thoughts about them, but in the context of the Cahiers, a project about the mind, they are of no account. What instead the Cahiers do offer is a demonstration of how to sustain a focused enquiry within which there was room for shifts and turns, for obliqueness as well as directness, and which was directed not towards a result, but towards the very act of thought itself. The Cahiers lay open a way of thinking in which orderliness and neatness are not necessarily virtues, and may indeed frustrate enquiry and originality. As Italo Calvino remarked of Valéry, as of Ortega y Gasset, ‘I read […] not to be convinced by his ideas but for the pleasure of seeing how the mechanism of his mind works’ (Calvino 323).