Art and Psychology: A Phenomenological Inquiry into the Nature of Pictorial Representations

Author(s): Luca Taddio (University of Udine) Year: 2025 (accepted 6 May 2025) Type: journal article (Author's contribution: solely L.T.)

A synoptic essay arguing that the central interpretive question of Merleau-Ponty's Eye and Mind — "What, then, is this secret science which he has or which he seeks?" — is most productively answered by perception, more specifically Gestalt psychology. Taddio proposes that the painter (paradigmatically Cézanne and Magritte) recreates the world by working implicitly with the modalities of phenomenal givenness that Gestalt experimental phenomenology has formalized — figure-ground, the laws of unification (Wertheimer), amodal completion (Kanizsa, Wittgenstein), conditions of transparency (Kanizsa), and depth cues (Gibson). On this reading, theories of pictorial representation grounded in resemblance (Gombrich, Hopkins, Wollheim) or denotation (Goodman) miss the explanatory layer at which an image becomes an image at all: the layer of phenomenal invariants — the conditions of appearance that operate identically in worldly perception and in pictorial perception. The essay weaves an Italian experimental-phenomenological tradition (Bozzi, Burigana, Massironi, Kanizsa) into Merleau-Pontyan ontology, sutured by a synthetic terminological core: système d'équivalences, modalities of appearance, phenomenal invariants.

Core Arguments

  1. Claim: Merleau-Ponty's "secret science" in Eye and Mind is perception, more specifically Gestalt psychology — the conditions of phenomenal givenness experimental phenomenology has formalized. Because: Cézanne creates "worlds" by drawing upon perceptual modalities underlying phenomenal appearance; Gestalt psychology has uncovered these rules through identification of dependent and independent variables. The painter's activity recreates the world using the same conditions of appearance the world employs. Against: Readings that treat MP's "secret science" as occult, mystical, or unspecifiable; against readings on which the painter-Gestaltist relation is mere parallelism.

  2. Claim: MP's relation to Gestalt psychology is bidirectional — he critiques Gestalt for retaining a naturalistic residue rooted in realism while reinterpreting Husserl in light of Gestalt's "new psychology." Because: The "naturalistic residue" — the Gestaltists' unexamined objectivism — blocks the move beyond the realism-idealism dichotomy. But Gestalt's experimental findings about conditions of appearance generate phenomenological content Husserlian phenomenology by itself does not. MP's dual movement: Husserl-against-Gestalt-residue + Gestalt-findings-against-Husserl-idealism. Against: Pure phenomenology of consciousness on one side; pure experimental psychology on the other.

  3. Claim: Pictorial representation is grounded not in resemblance (Gombrich, Hopkins, Wollheim) or denotation (Goodman) but in phenomenal invariants — the conditions of appearance shared by world and image. Because: (a) "Resemblance is the result of perception, not its mainspring" (MP); (b) Goodman's denotational/symbolic theory presupposes already-functional images that codes can interpret, but the conditions of image-being operate prior to coding; (c) Gestalt-identified phenomenal invariants (figure-ground, principles of unification, amodal completion factors, conditions of transparency, depth cues) operate identically in both world-perception and image-perception. The explanatory base is the phenomenal-invariant level, not the symbolic/denotational level. Against: Gombrich's "psychology of pictorial representation" (insofar as tied to learned schemas); Goodman's Languages of Art (denotation as foundational); resemblance theories more broadly.

  4. Claim: Painters do not imitate the world but recreate it. Resemblance, when it occurs, is a consequence of what painters achieve, not its foundation. Because: The painter, through their body — whose subjectivity is constituted by its relation with the world — engages the conditions of phenomenal appearance and produces a "second nature" (the painting itself) with its own internal conditions of givenness. The output is a système d'équivalences, not a copy. Against: Imitative theories of art; mimesis as foundational concept; correspondence theories of pictorial truth.

  5. Claim: World and pictorial representation share a système d'équivalences — same phenomenal organization governs both. The painting's depth, transparency, figure-ground, and amodal-completion operate on the same conditions as worldly depth, transparency, figure-ground, and amodal-completion. Because: Wertheimer's principles, Rubin's figure-ground, Kanizsa's amodal-completion factors, Kanizsa's three transparency conditions (topological + figural + chromatic), Gibson's monocular depth cues — all are observable in both world and painting. The painter "explores, experiments, and discovers worlds whose possibilities are already intrinsically contained within visual perception." Against: Any view treating painting as operating by a categorically different logic from perception (pure symbolism, pure illusion, pure denotation).

  6. Claim: Perception cannot be reduced to causal chains between distal stimulus, proximal stimulus, and brain. Because: The Kanizsa triangle (we see a triangle absent at distal-stimulus level); amodal completion (we see a single line passing behind an occluder where no continuous line exists physically); transparency (Magritte's L'abandon shows transparency on a physically opaque canvas); Köhler's stimulus error — confusing immediate experience with material substrate. The phenomenal world has properties (transparency, depth, color, expressivity) irreducible to physical layers. Against: Empiricist-associationist sensation-based explanations; representationalism; neuroreductive accounts.

  7. Claim: Depth is the primordial dimension — "the experience of the reversibility of dimensions" (MP) — and pictorial depth functions as a système d'équivalences with worldly depth. Because: Monocular depth cues (perspective, occlusion, texture gradients, elevation, relative size, shading) generate depth-impression in painting and depth-impression in the world via the same conditions. Pictorial depth is not representation of worldly depth but instance of the same dimension. Magritte's Golconde deploys texture gradient as such a system of equivalences. Against: Geometric/Cartesian space; depth as derived; depth as merely psychological cue.

  8. Claim: Art has a theological-metaphysical anthropological function (the mystery of existence, the punctum caecum) which the rise of science erodes; art and science are inversely proportional. Because: Art has historically functioned as bridge between subjectivity and the entirety of creation. Science's progress has redefined the boundaries of religion, art, philosophy. Neuroscience marks the loss of the artwork's "aura" (Benjamin). Were creative processes ever fully explained scientifically, art would lose its original anthropological function of expressing what reason cannot dominate (birth, death, the punctum caecum). Against: A purely cognitivist or instrumentalist view of art; reduction of aesthetic experience to neural processes.

Argumentative Movement

The essay moves through eight chapters organized as a cascading argument from MP's framing question (§1) through the Gestalt-experimental machinery (§§4–7) to the explicit displacement of resemblance/denotation theories (§8). The architectural logic: §1 frames the paper's interpretive thesis (secret science = Gestalt psychology) and the broader anthropology of art-vs-science. §2 introduces Kanizsa-anomalous-margins and Magritte's Le Jockey perdu / La condition humaine as case studies that dismantle physical-stimulus reductionism. §3 (amodal completion) generalizes the dismantling using Wittgenstein's animal/arrow example. §4 establishes figure-ground as precondition of phenomenal unity. §5 fills in Wertheimer's six laws. §6 adds transparency. §7 raises depth as primordial dimension and links Cézanne's "deflagration of Being" to the "system of equivalences." §8 ("Doing Things with Phenomena") concludes with the rejection of resemblance/denotation theories and the closing return to MP's "secret science" question.

The movement is not a deductive descent from a master thesis but an inductive ascent: each Gestalt principle is shown to operate on canvas as on world, until the cumulative weight authorizes the claim that the same set of phenomenal invariants operates across the two registers.

Key Findings

  • The central interpretive bid: MP's "secret science" (single-attestation E&M phrase) refers to perception understood as Gestalt psychology has formalized it, not to an occult or unspecifiable science.
  • A bidirectional reading of MP's relation to Gestalt: Husserl is used to critique Gestalt's naturalistic residue, while Gestalt findings are used to reinterpret Husserl beyond pure consciousness-philosophy.
  • A new explanatory base for picture theory: phenomenal invariants displace both resemblance (Gombrich/Hopkins/Wollheim) and denotation (Goodman) as the foundation of pictorial representation.
  • A unified account of pictorial perception via système d'équivalences: world and image share the same phenomenal-organizational rules; the painter recreates rather than imitates.
  • A re-anchoring of MP's depth in experimental-phenomenological terms: Cézanne's "deflagration of Being" can be analyzed via Gibson's monocular depth cues operating as a system of equivalences.
  • An anthropological framing: art's original function (the punctum caecum, the mystery of existence) is structurally threatened by scientific reduction, but Gestalt-grounded phenomenology preserves art's philosophical importance by treating its conditions as branches of Being (MP) rather than psychological epiphenomena.

Methodology

The paper combines: (a) philological-exegetical engagement with MP (via Eye and Mind, Phenomenology of Perception, Visible and the Invisible); (b) case-driven phenomenological description of specific paintings (Magritte's Le faux miroir, La condition humaine, Le Jockey perdu, Le Blanc-seing, L'idée, Le pèlerin, Paysage de Baucis, L'abandon, Golconde; Cézanne's Mont Sainte-Victoire; Escher's Eight Heads, Metamorphosis II); (c) systematic appropriation of Gestalt experimental-phenomenological findings (Wertheimer's laws, Rubin's figure-ground, Kanizsa's amodal-completion and transparency factors, Gibson's monocular depth cues); (d) reference to the Italian experimental-phenomenology tradition (Bozzi 1989; Burigana 1996; Massironi 1998; Kanizsa 1980, 1991) as the methodological apparatus making the bridge from psychology to ontology possible.

The "phenomenal invariants" terminology Taddio introduces for the relationship between dependent and independent variables — the conditions experimental phenomenology has identified as prerequisites of phenomenal appearance — is the technical lever connecting the experimental tradition to MP's late ontology.

Concepts Developed

  • phenomenal-invariants — Taddio's central technical contribution: the relationship between dependent and independent variables identified by experimental phenomenology as conditions for the appearance of a given phenomenon. Same set of invariants for representation and the thing represented. The painter works through invariants implicitly. Used to displace resemblance/denotation as foundation of picture theory. Primary site for this concept in the wiki.
  • systeme-d-equivalences — Already an MP figure (E&M §§ 2, 4) but Taddio's HUB-level deployment as the central explanatory device for the world-painting parallelism makes this paper a primary site. Spans §§ 1, 2, 7, 8 with 6+ attestations.
  • amodal-completion — Section-length treatment (§3) drawing on Kanizsa, Wittgenstein, Magritte's Le Blanc-seing. Primary site for this concept on the wiki.
  • gestalt-principles-of-unification — Section-length treatment (§5) systematizing Wertheimer's six laws with examples. Primary site.
  • figure-ground-relationship — Treated in §4 via Rubin and Koffka, with Escher's Eight Heads as illustration. Primary site.

Concepts Referenced

  • science-secrete — Taddio offers a competing reading (= perception/Gestalt psychology) of MP's question-figure. See Positions on the science-secrete page.
  • depth-profondeur — extended in §7 by reading pictorial depth via monocular cues as a system-of-equivalences with worldly depth.
  • fundamental-thought-in-art — converges with Taddio's reading of painting as discipline of access to phenomenal conditions.
  • chiasm, reversibility — used briefly in §8 ("the chiasm between our corporeality and the world").
  • flesh-as-element — used as MP's metaphor for phenomenal continuity (§§ 1, 6, 7, 8).
  • lebenswelt — used repeatedly in Husserlian sense.
  • pregnancy-pragnanz — implicit in §§ 5–7 wherever Gestalt's "good form" is invoked, but Taddio does not develop the concept; the register is experimental, not ontological-generative.
  • merleau-ponty-1961-eye-and-mind — primary text under reading throughout.
  • merleau-ponty-1945-phenomenology-of-perception — secondary anchor (esp. via the §2 "Each part arouses the expectation of more than it contains" passage).
  • merleau-ponty-1968-visible-and-invisible — citation locus for "It is hence because of depth that things have a flesh."
  • operational thinking (pensée opératoire) — implicit foil throughout via the anti-physicalist/anti-reductionist program.
  • affordance (Gibson) — used as cognate of expressive quality (§§ 5, 8).
  • Dasein (Heidegger) — brief grounding move §1.
  • body schema (MP) — used in §1.
  • world-image apparatus (Lorenz) — used twice as silent-key bridge between ecological theory and phenomenology.

Terminology

For technical-terminological tracking, the paper's English/Italian/French braid is loose — the paper is in English with Italian methodological apparatus and French (MP) primary anchors. Key technical terms:

Term Primary attribution Recurrence in the paper Role
système d'équivalences MP, but Taddio uses 6+ times §§ 1, 2, 7, 8 HUB explanatory term: world ↔ painting parallelism
phenomenal invariants Taddio's coinage (English) §§ 3, 6, 7, 8 Taddio's technical contribution
modalities of appearance / conditions of appearance Taddio (English) throughout Standing technical refrain
secret/silent science MP (E&M) §§ 1, 2, 8 MP-citation, structural framing
naturalistic residue MP (per Taddio) §1 only Silent-key — licenses dual-movement reading
stimulus error Köhler 1929 §6 only Silent-key — licenses anti-physicalism
branches of Being MP (E&M §4) §7 only Silent-key — licenses Gestalt-to-ontology bridge
punctum caecum Lacan-via-MP usage §1 only (multiple mentions) Limit-of-representation marker
amodal completion Kanizsa (1991) §3 (chapter title) Section-anchor
figure-ground Rubin (1915), Koffka §4 (chapter title) Section-anchor
principles of unification Wertheimer (1923) §5 Section-anchor cluster
transparency factors Kanizsa (1980) §6 Section-anchor
texture gradient / monocular depth cues Gibson (1986) §7 Section-anchor
affordance Gibson §§ 5, 8 Cognate-of-expressivity

Key Passages

"According to our working hypothesis, the 'secret science' to which Merleau-Ponty refers is perception, and more specifically, Gestalt psychology." (§1)

"Cézanne, in particular, creates 'worlds' by drawing upon the perceptual modalities underlying phenomenal appearance—rules that Gestalt psychology has experimentally uncovered through the identification of their conditions. The connection to this specific tradition lies in the phenomenological method employed by Gestalt psychologists, which focuses on 'functional relationships' (Bozzi, 1989: 27) and the identification of dependent and independent variables (Burigana, 1996: 274) that determine the conditions for the emergence of the phenomenon under investigation (Bozzi, 1989: 48)." (§1)

"Merleau-Ponty critiques Gestalt psychologists for retaining a naturalistic residue rooted in realism—an objectivist assumption that obstructs a proper understanding of the nature of phenomena. These unexamined biases, he argues, impede the possibility of transcending the realism-idealism dichotomy. Through his critical investigation within experimental psychology, Merleau-Ponty develops a novel interpretation of the phenomenological method characterized by a dual movement: on the one hand, he draws on Husserl to critique Gestalt psychology, while on the other, he reinterprets Husserl's work in light of the 'new psychology' (Taddio, 2024: 15)." (§1)

"Painters disclose 'things' that are the expression of a système d'équivalences secretly linked to that dimension of Being Merleau-Ponty describes as 'raw' or 'wild': worlds belonging to the same (phenomenal) 'fabric' – to use the author's metaphor – from which the entire sphere of the experienceable is woven." (§1)

"while working on the canvas, the painter plays with – or puts into play – these organizational forces, which Gestalt psychology has formalized. Cézanne's investigations reveal, 'by remaining faithful to the phenomena, what recent psychologists have come to formulate' (Merleau-Ponty, 1964a: 14)." (§3)

"We can describe the conditions of a thing's appearance on the canvas in terms of 'phenomenal invariants' that constitute the perceptual structure of the event (in the specific case of pictorial representation). These invariants represent both the dependent and independent variables underlying the perceived phenomenon. Within the tradition of experimental phenomenology, such relationships are directly observable properties." (§2 close)

"To deny transparency simply because we know the object is solid and opaque implies the 'stimulus error': confusing the plane of immediate experience with what we know about it (Köhler, 1929)." (§6)

"Pictorial art is a 'system of equivalences' internal to appearance; it is the phenomenon itself, together with the rules of its manifestation, that we share with the world. The meaning of this participation in the visible does not vanish within the image but emerges through it." (§7)

"By 'phenomenal invariants', we also mean what is common to both the representation and the thing represented. The painter works through these invariants, consciously or not, assessing the explicit result of their work as it takes shape." (§7)

"the conditions for the emergence of each factor – figure-ground, principles of unification, depth, and transparency – are, as Merleau-Ponty defines them, 'branches of Being' (Merleau-Ponty, 1964a, 1964b: 188) and can be grasped by identifying the variables that perceptually compose the structure of the event." (§7)

"An image becomes similar when the artist discovers expressive conditions akin to the object, conveying an invariant core of phenomenological origin." (§8)

"Resemblance is the result of perception, not its mainspring" (Merleau-Ponty: 171, cited §8)

"It is an image only as long as it does not resemble its object" (Merleau-Ponty, 1964a, 1964b: 170, cited §8)

"The nature of representation does not lie in resemblance, nor, as Goodman argued, in denotation. Rather, the representation of a thing replaces the thing itself in the sense that it recreates it: the image participates in the same rules of 'givenness' as the object." (§8)

"We can thus return to Eye and Mind and the question we posed initially: 'What, then, is this secret science which he has or which he seeks?' The answer lies in perception. Perception represents a significant point of intersection between art and science." (§8)

What's Not Obvious

  1. The paper's strongest move is its silent restructuring of picture theory's explanatory order. The dominant 20th-century anglophone theories (Gombrich's resemblance-and-schema; Goodman's denotation) take the question of pictorial representation to be: what relation must obtain between an image and what it depicts for the image to count as representing it? Taddio's move, channeling MP, is to displace this question entirely: representation is not a relation between two pre-existing terms (image and object) but the recreation of a phenomenal-organizational field shared by both. The displacement is performed not in §1 but in §7's definition of "phenomenal invariants" and §8's anti-Goodman/anti-Gombrich closing. This is the shift that makes the paper philosophically substantive rather than merely an MP-Gestalt synthesis. Notably, the displacement leaves the question of how conventional picture-codes (e.g., diagrams, maps, perspective notation) function — precisely Goodman's primary focus — without an immediate answer; the paper concedes this implicitly by restricting its examples to figurative painting where the world-painting système d'équivalences is most operative.

  2. "Secret science" is read against the wiki's existing reading. The wiki's science-secrete page (built mostly through the 2026-04-25 silent-key audit and Paper A's H_synth) reads MP's science secrète as a question-figure for indirect ontology as practiced through painting — undefined by design, with painting as primary witness. Taddio's reading is closer to a content-determination: the secret science is perception/Gestalt psychology. The two readings are not strictly incompatible — Taddio's content-determination could be a partial-determination of the H_synth question-figure (the painter's discipline of indirect access happens via the modalities Gestalt has formalized) — but the registers differ. A reader expecting either reading will find passages in MP that fit it; the truth is presumably that Eye and Mind is doing both: orienting via undefined question (H_synth) and gesturing toward Gestalt-experimental content (Taddio). This page records both as Positions on science-secrete without forcing adjudication.

  3. The §1 art-vs-science framing is in tension with the §§2–8 method. §1 develops a Heidegger-Severino-Damasio thesis: science erodes art's anthropological function (the mystery of existence at the punctum caecum); art and science are inversely proportional; were creative processes fully explained, art would lose its original aura. But §§2–8 ground a philosophical understanding of pictorial representation precisely in the experimental-scientific findings of Gestalt psychology. The tension: if scientific explanation of art's processes erodes art's aura, then explaining picture theory through Gestalt (a scientific psychology) seems to perform the very erosion §1 warns against. Taddio's tacit resolution — via MP's "naturalistic residue" critique — distinguishes Gestalt's findings (descriptive, can be retrieved into phenomenology) from Gestalt's frame (objectivist, must be critiqued). But the resolution is not theorized; it is left as the paper's working assumption. A reader could press on this point and the paper would have to be supplemented (probably from Taddio 2024's MP monograph) to answer.

Critique / Limitations

  • Internal tension between art-erosion-by-science framing (§1) and Gestalt-as-explanatory-base method (§§2–8). The bidirectional reading of MP's relation to Gestalt is offered as the resolution but is asserted rather than systematically defended.
  • Strong identification of pictorial and worldly invariants is contestable. There are conditions specific to pictorial perception (the framing of the canvas, conventional limits of pictorial space, durative engagement with a static image) and conditions specific to worldly perception (binocular disparity, motion parallax, temporal flow) that the paper does not address. The "système d'équivalences" thesis may be too strong as stated.
  • Goodman's denotation theory may not be refuted but talked past. Goodman operates at a level of conventional symbol-systems that Gestalt-level invariants do not directly address. Taddio's anti-denotation argument shows that figurative painting is not best explained by denotation; it does not show that all symbol-systems are best explained by phenomenal invariants. The paper's restriction to figurative cases (Cézanne, Magritte) is what makes the displacement work, and it does not engage Goodman's primary cases (diagrams, maps, perspectival notation).
  • The Gestalt tradition cited is canonical pre-1960. The post-Marr cognitive-science complications (Bayesian inference, predictive coding, dynamic systems) are not engaged. Wertheimer's laws of unification are post-1960 understood as defaults in particular contexts, not universal regulative laws. The paper's appropriation of Wertheimer is at the philosophical-phenomenological level rather than the experimental-robustness level — defensible, but not stated as such.
  • No engagement with Carbone, Saint Aubert, or the contemporary MP-aesthetics literature. The paper cites MP's own texts and the Italian experimental tradition but does not engage with the contemporary secondary literature on MP's aesthetics (Carbone's painting/cinema work; Saint Aubert's archive scholarship; Johnson's Poetic of the World). The reading of MP is consequently older-style and synoptic rather than nuanced by recent scholarship.
  • The Severino-Heidegger framing of §1 is applied without engagement. Severino's Essence of Nihilism and Destino della tecnica, Heidegger's Origin of the Work of Art, Lacan's Seminar XI, Cassirer, Lévi-Strauss, Damasio, Dunbar are all gestured to in §1 but none is engaged in argumentative depth. The framing is suggestive rather than substantive.

Connections

Sources

  • merleau-ponty-1961-eye-and-mind — primary object. Cited at every section. The "secret science" question (§1 close in E&M) is the paper's framing question.
  • merleau-ponty-1945-phenomenology-of-perception — secondary anchor for the anti-empiricism program; the "Each part arouses the expectation of more than it contains" passage is loadbearing in Taddio §2.
  • merleau-ponty-1968-visible-and-invisible — citation locus for "It is hence because of depth that things have a flesh" (V&I 219), cited Taddio §7.
  • External non-wiki sources cited extensively: Bozzi 1989 Fenomenologia sperimentale; Burigana 1996 Singularità della visione; Massironi 1998 Fenomenologia della percezione visiva; Kanizsa 1980 Grammatica del vedere, 1991 Vedere e pensare; Koffka 2013 Principles of Gestalt Psychology (1935); Köhler 1929 Gestalt Psychology; Wertheimer 1923 (English in Ellis 1938); Gibson 1986 The Ecological Approach to Visual Perception; Gombrich 1960 Art and Illusion; Goodman 1976 Languages of Art; Wittgenstein 1980 Remarks on the Philosophy of Psychology, 2003 Philosophical Investigations; Putnam 1981 Reason, Truth, and History; Lacan 1977 Seminar XI; Heidegger 2004 Being and Time, 2008 Origin of the Work of Art; Magritte 2016 Selected Writings; Severino 2016 Essence of Nihilism, 2009 Il destino della tecnica; Simondon 2020; Tagliagambe 2020.
  • Self-references: Taddio 2019 (Wittgenstein and aesthetics); Taddio 2023 (digital phenomenology, Foundations of Science); Taddio 2024 (Merleau-Ponty monograph, Feltrinelli — cited as the principal source for the bidirectional reading of MP's relation to Gestalt).