Pith. sign in

REVIEW 4 major objections 6 minor 3 cited by

Why does time feel the way it does? Towards a principled account of temporal experience

T0 review · 4 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper argues that the feeling of time flowing is explained by the cause-effect structure of directed grids in the brain, making time a structure rather than a process.

desk verdict IIT's directed-grid story for temporal flow is a genuine extension of the spatial account, but the universal claim outruns the demonstrated examples. read the letter →

arxiv 2412.13198 v1 pith:74SJC7CI submitted 2024-11-30 q-bio.NC

Renzo Comolatti , Matteo Grasso , Giulio Tononi This is my paper · ORCID
classification q-bio.NC
keywords temporalexperienceintegratedinformationtheorycause-effectstructuredirectedgridspeciouspresentflowconsciousnessIIT4.0
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper tries to give an objective account of why time feels flowing, not by appealing to clocks or to a process unfolding in the brain, but by showing that a specific causal structure can reproduce the felt order of moments. It claims that four introspectible properties of temporal experience—directedness, directed inclusion, directed connection, and directed fusion—are necessary and sufficient for the feeling of flow, and that these correspond exactly to properties of the cause-effect structure unfolded from a directed 1D grid under Integrated Information Theory 4.0. If correct, the conscious present is a single static cause-effect structure specified by the brain's current state, so time is a structure rather than a process. A sympathetic reader would care because this would extend a principled physical account of the quality of experience from space to time, with testable predictions about the duration and grain of the present.

What carries the argument

The load-bearing object is the directed 1D grid: seven binary units with asymmetric nearest-neighbor connections (self-connection weight 0.3, stronger lateral weight 0.6, weaker lateral weight 0.1), driven by an input array that acts as a delay line preserving a trace of roughly 210 milliseconds of clock time. Under the IIT 4.0 formalism, the grid's current macro state (about 30 milliseconds) unfolds into a cause-effect structure of causal distinctions—mechanisms with cause and effect purviews—and causal relations formed by congruent overlaps among purviews. The asymmetry between cause and effect purviews gives directedness, and the systematic overlaps among distinctions realize directed inclusion, connection, and fusion; a cause-effect structure satisfying all of these is called a flow.

What would settle it

A psychophysical test could settle the sufficiency claim: if subjects can report differences in the experienced speed or felt passage of time while the four structural properties are held constant, then the four properties are not sufficient and the directed-grid flow would not explain the feeling of temporal flow.

Watch

Extended reading notes

Core claim

The paper's central claim is that a directed 1D grid of seven probabilistic units, in its current macro state, unfolds under IIT 4.0 into a cause-effect structure in which causal distinctions are directed—causes lean toward the now and effects toward the then—and are bound by relations of directed inclusion, connection, and fusion. These structural properties match, one by one, the phenomenal properties that the paper argues make up the experience of temporal flow: moments point away from themselves, nest within one another, overlap in an ordered way, and compose larger moments. The experienced present is therefore not a process unrolling in clock time; it is a cause-effect structure specified by a system in its current state. Time, on this account, is a structure, not a process.

Load-bearing premise

The argument rests on the introspective claim that the feeling of temporal flow is fully captured by four properties—directedness, directed inclusion, directed connection, and directed fusion—so if flow has an essential ingredient beyond these, such as experienced speed or a separate sense of passage, the directed-grid account would not explain it.

Editorial extensions

If this is right

  • If the account is right, the felt present is not a replay of neural activity over hundreds of milliseconds; it is a single macro state lasting about 30 milliseconds that encodes a roughly 210-millisecond trace through its cause-effect structure.
  • The duration of the experienced present should scale with the number of units in a directed grid, so larger grids would support longer presents and higher-level areas could bind longer sequences of events.
  • Artificial activation near the now terminus of a directed grid should make a stimulus feel present, while activation near the then terminus should make it feel as if it occurred earlier.
  • Postdictive effects, protention, and the felt salience of the now are accommodated by top-down influences and by denser connectivity at the now end of the grid.
  • Temporal and spatial experience share the same explanatory scheme: directed 1D grids yield flow, while non-directed 2D or 3D grids yield spatial extension.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the four properties are truly sufficient, then any physical system whose unfolded cause-effect structure is a flow—not just neural grids—would have temporal experience; this is an implication the paper does not develop.
  • The account suggests that experienced duration, boundary, and interval are derived from inclusion and connection relations rather than read off clock time, which could give a structural explanation of time illusions such as the oddball effect or temporal ventriloquism.
  • The claim that time is a structure rather than a process, if extended, would align with block-universe views: passage would be an intrinsic property of a state rather than motion along a temporal axis, which would reframe, not resolve, the metaphysics of time.
  • One testable extension is that if the now terminus is preferentially connected to action-related areas, then perturbing those connections should shift the felt now relative to action without changing the felt order of moments.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The paper proposes an IIT 4.0 account of temporal experience. It argues that the conscious present is experienced as flowing because it is composed of directed phenomenal distinctions (moments) bound by directed inclusion, directed connection, and directed fusion. The authors then model a seven-unit directed 1D grid with an external input delay line, unfold its cause–effect structure using PyPhi, and show that several contiguous distinctions (bcDe, cDef, bcDef, cDe) and their 2-relations exhibit properties corresponding to the four phenomenal properties. They conclude that the experienced present does not correspond to a process in clock time but to a static cause–effect structure, so that time is a structure rather than a process.

Significance. If the full cause–effect structure of directed grids indeed satisfies the four flow properties, the paper extends IIT's explanatory identity from spatial extension to temporal flow, offering a principled and computationally explicit account of why time feels flowing. The use of PyPhi makes the worked computations reproducible, and Section 4.13 contains falsifiable predictions (e.g., present duration proportional to grid size, now/then terminus activation effects). However, the universal claim currently rests on a partial analysis of the cause–effect structure and on a strong phenomenological premise about the sufficiency of the four properties. The paper is transparent about several of its limitations, which is commendable, but the gap between the demonstrated examples and the global conclusion must be addressed before the central identity claim is fully supported.

major comments (4)
  1. [§3.1.2, §4.1, Figure 8 caption] The paper's central claim—that the cause–effect structure of a directed grid is a flow—is demonstrated only for a subset of distinctions and relation types. Section 3.1.2 explicitly limits the analysis to 2-relations and 2-faces, and Section 4.1 states that 'nearly a third of the distinctions are specified by contiguous units' and that these will be the focus. The Figure 8 caption then asserts, without computation or proof, that 'This holds for the other distinctions that compose the cause–effect structure.' Since the paper's own phenomenology (Section 2.1) includes moments such as 'two notes separated by a pause,' which are non-contiguous, and since higher-degree relations are part of the cause–effect structure, the universal identity claim is not supported by the presented evidence. This gap is load-bearing for the Abstract's conclusion that directed grids account for the feeling of time flowing.
  2. [§4.2–4.5, §3.2] The computational demonstration is a single instance: a 7-unit directed grid with hand-chosen weights (self-weight 0.3, lateral weights 0.6 and 0.1) in the single state abcDefG. The paper does not show that directedness, directed inclusion, directed connection, and directed fusion hold for other weight configurations, other states, or other grid lengths, nor does it provide an analytic argument for arbitrary directed grids. Without such a generalization, the statement in §4.6 that 'the cause–effect structure unfolded from a directed grid is composed of distinctions that are directed...' is an overgeneralization from one example.
  3. [§2, §4.6, §4.13] The sufficiency of the four phenomenal properties is asserted rather than established. Section 2 states that directedness, directed inclusion, directed connection, and directed fusion are 'necessary and sufficient for the experience of time,' but no independent operationalization or empirical criterion is given; 'flow' is then defined in §4.6 simply as a cause–effect structure satisfying those same properties. This makes the central correspondence partly circular. The paper also sets aside experienced speed in §2 ('the feeling of temporal flow only refers to... rather than to the experienced speed'), yet §4.13 later appeals to modulations of synaptic strength to account for slowing or speeding of time; the authors should clarify how these statements are compatible.
  4. [§4.12] The conjecture that arrays of directed grids in auditory and other posterior cortices are the substrate of temporal experience is acknowledged to have little direct evidence. This is a legitimate hypothesis, but it means the predictions in §4.13 are conditional on an unverified substrate. The paper should more sharply separate the formal cause–effect result—what directed grids would imply if they were the substrate—from the empirical conjecture about where such grids are located in the brain.
minor comments (6)
  1. [§3.2] The sentence 'The input array not only provides bottom-up inputs...; is also works as a delay line' contains a typo: 'is also works' should be 'it also works.'
  2. [§3.2] The displayed equation for sigma(I*; s_k, w_k, I) is typeset with garbled subscripts and superscripts; it needs proper mathematical formatting for the equation to be readable.
  3. [References] The text cites Arstila (2015) in §4.13 and Arstila (2018) in §4.15, but the reference list contains only Arstila (2023); please add the missing references or correct the in-text citations.
  4. [Figure 8 caption] In the caption, 'coincides with the union the purviews of bcDef' is missing 'of': it should read 'coincides with the union of the purviews of bcDef.'
  5. [§4.12] The phrase 'Lateral and back-connections connections' contains a duplicated 'connections'; it should be 'Lateral and back-connections.'
  6. [§3] The PyPhi citation appears as 'W. G. P. Mayner et al., 2018' in the text but as 'Mayner, W., Marshall, W., ...' in the reference list; please standardize the citation style.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the flow properties are computed consequences of an explicit directed-grid substrate, while reliance on IIT is a stated theoretical framework rather than a self-citation-based reduction.

full rationale

The paper's derivation chain is not circular. The substrate is explicitly specified (seven-unit directed grid abcDefG with stated connection weights and activation function, Section 3.2), and the cause–effect structure is unfolded by PyPhi, so the properties of directedness, directed inclusion, connection, and fusion are computed properties of that structure rather than assumptions inserted into the conclusion. The phenomenal properties in Section 2 are characterized independently by introspection and are not defined in terms of cause–effect structure; the correspondences in Sections 4.2–4.5 are substantive mappings between two separately described domains. The paper does rely on IIT's explanatory identity (cited to Albantakis et al. 2023) and on prior work by the same authors for the spatial analogy and the subtext/supertext notions, but these are framework premises, not results obtained by circular reduction. The main limitations are non-circular: Section 4.1 restricts the detailed analysis to contiguous distinctions ('nearly a third of the distinctions ... will be the focus for the account below'), and the Figure 8 caption asserts without proof that 'This holds for the other distinctions that compose the cause–effect structure'; Section 4.12 concedes that 'little is known about the presence and location of such directed grids in the brain.' These are gaps in evidence or proof, not cases where a prediction is equivalent to its input by construction. The directedness of the distinctions follows from choosing a directed grid, but that is an explicit modeling choice for the substrate, not a fitted parameter renamed as a prediction.

Assumptions & free parameters 6 free parameters · 4 assumptions · 2 invented entities

The central claim rests on the IIT identity axiom, an introspective phenomenological classification, a conjectured brain substrate, and hand-chosen model parameters. No independent empirical data are used; the PyPhi computation is theory-internal and is not calibrated against external measurements.

free parameters (6)
  • self-connection weight = w = 0.3
    Hand-chosen in Section 3.2; affects the transition probability matrix and which mechanisms are irreducible; no sensitivity analysis provided.
  • strong outgoing lateral weight = w = 0.6
    Hand-chosen in Section 3.2; establishes the dominant directed asymmetry of the grid.
  • weak outgoing lateral weight = w = 0.1
    Hand-chosen in Section 3.2; creates the weaker opposite direction in the directed grid.
  • macro state duration = ~30 ms
    Assumed in Section 4.8 as the intrinsic update grain; IIT says this should maximize phi_s, but no computation is shown for this substrate.
  • number of grid units = 7
    Chosen for tractability (Section 4.16); determines the modeled ~210 ms conscious present; no scaling analysis is provided.
  • illustrated current state = abcDefG (D and G ON)
    Selected for the worked examples in Sections 4.2 to 4.6; the cause-effect structure is state-dependent, so the demonstration is tied to this state.
assumptions (4)
  • domain assumption IIT explanatory identity: a cause-effect structure specified by a substrate in its current state fully accounts for the presence and quality of consciousness.
    Invoked in the Introduction and Section 3.1 as the theoretical foundation; not established by this paper.
  • domain assumption The four phenomenal properties (directedness, directed inclusion, directed connection, directed fusion) are necessary and sufficient for temporal experience.
    Introduced in Section 2 as introspective phenomenology; the paper states they are sufficient and uses them as the target for the physical correspondence.
  • ad hoc to paper The substrate of temporal experience is arrays of directed 1D grids in the brain.
    Conjectured in the Abstract and Section 4.12; the paper admits little is known about the presence and location of such grids.
  • domain assumption PyPhi and the IIT 4.0 formalism correctly compute causal distinctions, relations, and cause-effect structures.
    The Methods rely on PyPhi (Mayner et al., 2018) and IIT 4.0 (Albantakis et al., 2023) without independent verification in this paper.
invented entities (2)
  • directed grids (arrays of directed 1D grids) in auditory cortex
    purpose: Postulated neural substrate whose cause-effect structure is claimed to account for temporal flow.
    Section 4.12 states that little is known about the presence and location of such directed grids in the brain; no independent neurophysiological evidence is provided.
  • input array delay line
    purpose: Model component that drives the grid and preserves a ~210 ms trace of sensory input.
    Introduced in Section 3.2 as part of the model; not claimed to be a known anatomical structure, and no independent evidence is offered.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Why does time feel the way it does? Towards a principled account of temporal experience." pith.science (2026). https://pith.science/paper/74SJC7CI

@misc{pith2026241213198,
  author       = {Pith},
  title        = {Pith review of: Why does time feel the way it does? Towards a principled account of temporal experience},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/74SJC7CI}},
  note         = {Machine review of arXiv:2412.13198}
}
read the original abstract

Time flows, or at least the time of our experience does. Can we provide an objective account of why experience, confined to the short window of the conscious present, encompasses a succession of moments that slip away from now to then--an account of why time feels flowing? Integrated Information Theory (IIT) aims to account for both the presence and quality of consciousness in objective, physical terms. Given a substrate's architecture and current state, the formalism of IIT allows one to unfold the cause-effect power of the substrate, yielding a cause-effect structure. According to IIT, this accounts in full for the presence and quality of experience, without any additional ingredients. In previous work, we showed how unfolding the cause-effect structure of non-directed grids, like those found in many posterior cortical areas, can account for the way space feels--namely, extended. Here we show that unfolding the cause-effect structure of directed grids can account for how time feels--namely, flowing. First, we argue that the conscious present is experienced as flowing because it is composed of phenomenal distinctions (moments) that are directed, and these distinctions are related in a way that satisfies directed inclusion, connection, and fusion. We then show that directed grids, which we conjecture constitute the substrate of temporal experience, yield a cause-effect structure that accounts for these and other properties of temporal experience. In this account, the experienced present does not correspond to a process unrolling in "clock time," but to a cause-effect structure specified by a system in its current state: time is a structure, not a process. We conclude by outlining similarities and differences between the experience of time and space, and some implications for the neuroscience, psychophysics, and philosophy of time.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 3 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Intrinsic meaning, perception, and matching

    q-bio.NC 2024-12 conditional novelty 6.0 of 10

    Within IIT, perception is formalized as a stimulus-triggered portion of a Φ-structure, and perceptual differentiation and matching quantify how much intrinsic meaning an environment triggers.

  2. Dissociating Artificial Intelligence from Artificial Consciousness

    cs.AI 2024-12 conditional novelty 6.0 of 10

    Under Integrated Information Theory, functional equivalence between a digital computer and a simulated target does not imply phenomenal equivalence, so AI without consciousness is possible.

  3. On the utility of toy models for theories of consciousness

    q-bio.NC 2025-07 conditional novelty 3.0 of 10

    A consciousness researcher makes the case that toy models help clarify, test, and compare theories of consciousness, using IIT and GWT as case studies.

Reference graph

Works this paper leans on

17 extracted references · 11 canonical work pages · cited by 3 Pith papers

  1. [1]

    building blocks

    structure. 4.3. Directed inclusion Phenomenally, directed inclusion captures the fact that every moment includes and is included by other moments in a directed way, both towards the now (forward inclusion) and towards the then (backward inclusion). Figure 5: Directedness. Phenomenal distinctions in temporal experience are called moments. Moments are funda...

    work page 2019

  2. [3]

    https://doi.org/10.33735/phimisci.2022.9438 Singhal, I., & Srinivasan, N. (2021). Time and time again: a multi-scale hierarchical framework for time-consciousness and timing of cognition. Neuroscience of Consciousness, 2021(2), niab020. https://doi.org/10.1093/nc/niab020 Singhal, I., & Srinivasan, N. (2024). Just one moment: Unifying theories of conscious...

    arXiv 2021

  3. [5]

    J., Riedner, B., Boly, M., Postle, B

    https://doi.org/10.3389/fpsyg.2014.00196 Siclari, F., Baird, B., Perogamvros, L., Bernardi, G., LaRocque, J. J., Riedner, B., Boly, M., Postle, B. R., & Tononi, G. (2017). The neural correlates of dreaming. Nature Neuroscience, 20(6), 872-878. https://doi.org/10.1038/nn.4545 Singhal, I., Mudumba, R., & Srinivasan, N. (2022). In search of lost time: Integr...

    work page arXiv 2017

  4. [6]

    represents

    as well as cognitive maps (O’Keefe & Nadel, 1978). However, the anatomical organization of the hippocampal formation is very different from that of posterior cortex, making it less likely to be part of the substrate of consciousness. Lesion data also indicate that while the hippocampal formation is critical for supporting functions such as episodic memory...

    work page 1978

  5. [8]

    out there,

    or of how experienced time maps and represents clock time (Herzog et al., 2016, 2020; Hogendoorn, 2022; Northoff & Zilio, 2022). Instead, it is the one-to-one correspondence between the subjective, phenomenal properties of temporal experiences and objective, physical properties of the cause–effect structure unfolded from a certain kind of substrate. 4.15....

    arXiv 1998

  6. [41]

    subjective present

    Vuust, P., Heggli, O. A., Friston, K. J., & Kringelbach, M. L. (2022). Music in the brain. Nature Reviews Neuroscience, 23(5), 287-305. https://doi.org/10.1038/s41583-022-00578-5 Wackermann, J., Wittmann, M., Hasler, F., & Vollenweider, F. X. (2008). Effects of varied doses of psilocybin on time interval reproduction in human subjects. Neuroscience Letter...

    doi:10.1038/s41583-022-00578-5 2022

  7. [42]

    https://doi.org/10.1186/1471-2202-5-42 Tononi, G. (2008). Consciousness as Integrated Information: a Provisional Manifesto. The Biological Bulletin, 215(3), 216-242. https://doi.org/10.2307/25470707 26 Tononi, G. (2014). Why Scott should stare at a blank wall and reconsider (or, the conscious grid). Shtetl-Optimized. Tooley, M. (1997). Time, Tense, and Ca...

    doi:10.1186/1471-2202-5-42 2008

  8. [131]

    https://doi.org/10.1016/j.conb.2008.06.002 Eichenbaum, H. (2014). Time cells in the hippocampus: a new dimension for mapping memories. Nature Reviews Neuroscience, 23 15(11), 732-744. https://doi.org/10.1038/nrn3827 Ellia, F., Hendren, J., Grasso, M., Kozma, C., Mindt, G., Lang, J., Haun, A., Albantakis, L., Boly, M., & Tononi, G. (2021). Consciousness an...

    work page doi:10.1016/j.conb.2008.06.002 2014

Show all 17 references
  1. [334]

    S., & Meck, W

    https://doi.org/10.3390/e25020334 Matell, M. S., & Meck, W. H. (2000). Neuropsychological mechanisms of interval timing behavior. BioEssays, 22(1), 94-103. https://doi.org/10.1002/(SICI)1521-1878(200001)22:1 Mayner, W., Marshall, W., Albantakis, L., Findlay, G., Marchman, R., ...

    2000 doi

  2. [1160]

    J., & Ivry, R

    https://doi.org/10.3390/e21121160 Hayashi, M. J., & Ivry, R. B. (2020). Duration-selectivity in right parietal cortex reflects the subjective experience of time. The Journal of Neuroscience, JN-RM-0078-0020. https://doi.org/10.1523/JNEUROSCI.0078-20.2020 Hayashi, M. J., van de...

    2020 doi

  3. [1905]

    M., Hugueville, L., Lehéricy, S., Poline, J

    https://doi.org/10.3389/fpsyg.2015.01905 Ferrandez, A. M., Hugueville, L., Lehéricy, S., Poline, J. B., Marsault, C., & Pouthas, V. (2003). Basal ganglia and supplementary motor area subtend duration perception: an fMRI study. NeuroImage, 19(4), 1532-1544. https://doi.org/10.1...

    2003

  4. [2009]

    organs of succession,

    (see Dainton (2023) and White (2017) for critical reviews). The duration of instants, also challenging to introspect, has been estimated indirectly by assessing temporal order thresholds (the shortest inter-stimulus interval under which two sequential stimuli are perceived as ...

    2023

  5. [2013]

    imagining

    as well as for desired effects in music (Huron, 2006; Vuust et al., 2022). Top-down connections may also play a role in our capacity to hold specific contents, such as a sequence of numbers, within the experienced present, subserving functions such as working memory. Finally, ...

    2006

  6. [2014]

    postdictive

    in which stimuli occurring later can affect the experience triggered by stimuli occurring earlier. Such “postdictive” effects can be naturally accommodated within the present framework. For example, top-down connections from higher-level areas may affect the activation of unit...

    2008

  7. [2022]

    representation

    and employs these representations for motor control, memory, and cognitive functions. However, the framework presented here differs from cognitive and computational paradigms both with respect to what it tries to explain (the explanandum) and to how it tries to do so (the expl...

    2021

  8. [2023]

    static,” rather than to a process that actually “flows

    between the properties of temporal experience and those of the flow structure specified by directed grids. 4.8. Temporal flow as a directed structure A central aspect of IIT’s account is that the experience of time flowing corresponds to a directed structure that is “static,” ...

    2023

  9. [2024]

    McTaggart, J. E. (1908). The Unreality of Time. Mind, New Series, 17(68), 457-474. Merchant, H., Harrington, D. L., & Meck, W. H. (2013). Neural Basis of the Perception and Estimation of Time. Annual Review of Neuroscience, 36(Volume 36, 2013), 313-336. https://doi.org/10.1146...

    1908

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.