Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 46 of the 46 references it cites were found in a registry or an index.

Black-boxing and cause-effect power

Open access
Marshall, William, Albantakis, Larissa, Tononi, Giulio · 2018 · PLOS Computational Biology

Abstract

Reductionism assumes that causation in the physical world occurs at the micro level, excluding the emergence of macro-level causation. We challenge this reductionist assumption by employing a principled, well-defined measure of intrinsic cause-effect power-integrated information (Φ), and showing that, according to this measure, it is possible for a macro level to "beat" the micro level. Simple systems were evaluated for Φ across different spatial and temporal scales by systematically considering all possible black boxes. These are macro elements that consist of one or more micro elements over one or more micro updates. Cause-effect power was evaluated based on the inputs and outputs of the black boxes, ignoring the internal micro elements that support their input-output function. We show how black-box elements can have more common inputs and outputs than the corresponding micro elements, revealing the emergence of high-order mechanisms and joint constraints that are not apparent at the micro level. As a consequence, a macro, black-box system can have higher Φ than its micro constituents by having more mechanisms (higher composition) that are more interconnected (higher integration). We also show that, for a given micro system, one can identify local maxima of Φ across several spatiotemporal scales. The framework is demonstrated on a simple biological system, the Boolean network model of the fission-yeast cell-cycle, for which we identify stable local maxima during the course of its simulated biological function. These local maxima correspond to macro levels of organization at which emergent cause-effect properties of physical systems come into focus, and provide a natural vantage point for scientific inquiries.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1371/journal.pcbi.1006114

Cites

46 of 46 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig 1 Prototypical examples of black-box and coarse-grain elements that can lead to the emergence of macro-level cause-effect power. Left: A black-box element conceals many micro elements with specific functions. Right: A coarse grain macro element averages together many homogenous micro elements that share a global function.
Fig 1 Prototypical examples of black-box and coarse-grain elements that can lead to the emergence of macro-level cause-effect power. Left: A black-box element conceals many micro elements with specific functions. Right: A coarse grain macro element averages together many homogenous micro elements that share a global function.CC BY · PubMed Central
Fig 2 A schematic neuron considered as a number of `micro' elements (left), or as a black box (right). At the micro scale, the neuron receives inputs at its synapses (S), which are passed on to the cell body (C) and then to the axon hillock (A), which outputs to other neurons. Cause-effect power is assessed by perturbing each element (small hands) and observing the effects, while irreducibility is assessed by partitioning the elements (dashed red line). At the macro scale, there is only the black-box element (neuron) which receives three inputs and generates an output. Cause-effect power is assessed by perturbing the output of the black box (big hand) and observing its effects without constraining the constituent micro elements, however its irreducibility is still assessed by partitioning between micro elements (dashed red line).
Fig 2 A schematic neuron considered as a number of `micro' elements (left), or as a black box (right). At the micro scale, the neuron receives inputs at its synapses (S), which are passed on to the cell body (C) and then to the axon hillock (A), which outputs to other neurons. Cause-effect power is assessed by perturbing each element (small hands) and observing the effects, while irreducibility is assessed by partitioning the elements (dashed red line). At the macro scale, there is only the black-box element (neuron) which receives three inputs and generates an output. Cause-effect power is assessed by perturbing the output of the black box (big hand) and observing its effects without constraining the constituent micro elements, however its irreducibility is still assessed by partitioning between micro elements (dashed red line).CC BY · PubMed Central
Fig 3 An XOR logic gate with one time-step propagation delay. Left: Two COPY elements which each take a single input and relay the values to an XOR element. Middle: A black-box element with two inputs and a single output element (dashed red outline). Right: By perturbing the inputs to the black-box element in all possible ways, it is determined that it implements XOR logic over two time steps.
Fig 3 An XOR logic gate with one time-step propagation delay. Left: Two COPY elements which each take a single input and relay the values to an XOR element. Middle: A black-box element with two inputs and a single output element (dashed red outline). Right: By perturbing the inputs to the black-box element in all possible ways, it is determined that it implements XOR logic over two time steps.CC BY · PubMed Central
Fig 4 A system of three interconnected XOR elements with one-step propagation delay and all elements in the OFF state. Top: At the micro level, the propagation delay is modeled by COPY elements between the XOR elements (left). The micro level has only first-order mechanisms specified by individual XOR elements (right), and Φ = 0.25. Bottom: When the elements are black-boxed over two time steps, the system is comprised of three interconnected macro elements implementing XOR logic. The macro-level system has second-order mechanisms specified by pairs of XOR elements (right) and Φ = 1.875.
Fig 4 A system of three interconnected XOR elements with one-step propagation delay and all elements in the OFF state. Top: At the micro level, the propagation delay is modeled by COPY elements between the XOR elements (left). The micro level has only first-order mechanisms specified by individual XOR elements (right), and Φ = 0.25. Bottom: When the elements are black-boxed over two time steps, the system is comprised of three interconnected macro elements implementing XOR logic. The macro-level system has second-order mechanisms specified by pairs of XOR elements (right) and Φ = 1.875.CC BY · PubMed Central
Fig 5 Two potential mechanisms from the propagation delay network in Fig 4 which end up being reducible. On top is a COPY element that does not specify a mechanism. By being OFF in the current state, the COPY element constrains its input to be OFF in the previous state, but it does not constrain the future state of its output element, because the state of the XOR element still completely depends on the unknown state of its other input (shown here in grey). The bottom panel is a set of COPY elements which do not specify a high-order mechanism because they do not have an irreducible cause (the red line partitions the cause in two with no loss of information). Taking each COPY element independently fully constrains the past state of its input to be OFF.
Fig 5 Two potential mechanisms from the propagation delay network in Fig 4 which end up being reducible. On top is a COPY element that does not specify a mechanism. By being OFF in the current state, the COPY element constrains its input to be OFF in the previous state, but it does not constrain the future state of its output element, because the state of the XOR element still completely depends on the unknown state of its other input (shown here in grey). The bottom panel is a set of COPY elements which do not specify a high-order mechanism because they do not have an irreducible cause (the red line partitions the cause in two with no loss of information). Taking each COPY element independently fully constrains the past state of its input to be OFF.CC BY · PubMed Central
Fig 6 Left: A collection of 11 NOR elements at the micro scale. Middle: A macro scale black-boxing of these elements into four black boxes. Perturbing the inputs of these black boxes reveals that three implement AND logic and the final one OR logic, each at a time scale of two time steps. Right: A macro scale black-boxing with one element, implementing MAJORITY logic over its three inputs at a time scale of four time steps. Note that there are only three inputs, but each input arrives at two different micro elements.
Fig 6 Left: A collection of 11 NOR elements at the micro scale. Middle: A macro scale black-boxing of these elements into four black boxes. Perturbing the inputs of these black boxes reveals that three implement AND logic and the final one OR logic, each at a time scale of two time steps. Right: A macro scale black-boxing with one element, implementing MAJORITY logic over its three inputs at a time scale of four time steps. Note that there are only three inputs, but each input arrives at two different micro elements.CC BY · PubMed Central
Fig 7 A system of 55 interconnected NOR micro elements viewed at several different grain sizes, with all elements in the ON state. The micro elements form 5 interconnected groups, with each group arranged according to Fig 6. The systems are arranged with micro elements on the far left and black-box elements of increasing spatial grain to the right. The legend on the right specifies the input-output function of each element. Each of the three systems on the top row is a local maximum of cause-effect power, corresponding to the three spatiotemporal grains shown in Fig 6. Shown in the bottom row are two representative examples of the many systems with Φ = 0 at spatial grains between the local maxima.
Fig 7 A system of 55 interconnected NOR micro elements viewed at several different grain sizes, with all elements in the ON state. The micro elements form 5 interconnected groups, with each group arranged according to Fig 6. The systems are arranged with micro elements on the far left and black-box elements of increasing spatial grain to the right. The legend on the right specifies the input-output function of each element. Each of the three systems on the top row is a local maximum of cause-effect power, corresponding to the three spatiotemporal grains shown in Fig 6. Shown in the bottom row are two representative examples of the many systems with Φ = 0 at spatial grains between the local maxima.CC BY · PubMed Central
Fig 8 Boolean network model of the fission-yeast cell-cycle. (A) A network of 9 linear threshold elements connected by excitatory (black) and inhibitory (red) connections. The element, SK receives no input from the other elements. For this reason, we do not consider it as a part of the cell-cycle network, but rather as an external input that serves as a catalyst to initiate cell division. At the micro level, the system comprised of eight elements (in blue) is a stable local maximum of Φ for the duration of the biological sequence. (B) A sequence of 9 states that represent the cell division cycle, called the networks biological sequence.
Fig 8 Boolean network model of the fission-yeast cell-cycle. (A) A network of 9 linear threshold elements connected by excitatory (black) and inhibitory (red) connections. The element, SK receives no input from the other elements. For this reason, we do not consider it as a part of the cell-cycle network, but rather as an external input that serves as a catalyst to initiate cell division. At the micro level, the system comprised of eight elements (in blue) is a stable local maximum of Φ for the duration of the biological sequence. (B) A sequence of 9 states that represent the cell division cycle, called the networks biological sequence.CC BY · PubMed Central
Fig 9 All stable local maxima of macro cause-effect power for the cell-cycle network over the course of its biological sequence. Stable local maxima are identified at two different time scales (over 3 or 4 micro updates) and with groupings of the eight micro elements into either two or three macro elements. The output element for each black box is marked by a green outline; one common feature among all of the stable maxima is that element Slp1 acts as an output element of one black box. Note that connections between black boxes that do not originate from output elements are not shown in the figure because they do not contribute to the cause-effect structure (see S3 Text).
Fig 9 All stable local maxima of macro cause-effect power for the cell-cycle network over the course of its biological sequence. Stable local maxima are identified at two different time scales (over 3 or 4 micro updates) and with groupings of the eight micro elements into either two or three macro elements. The output element for each black box is marked by a green outline; one common feature among all of the stable maxima is that element Slp1 acts as an output element of one black box. Note that connections between black boxes that do not originate from output elements are not shown in the figure because they do not contribute to the cause-effect structure (see S3 Text).CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 46 references

Cited by

See all 81

Metadata

Authors
Marshall, W., Albantakis, L., & Tononi, G.
Affiliations
University of Wisconsin–Madison (US)
Corresponding author
Tononi, Giulio
Venue
PLOS Computational Biology
Published
2018-04-23
Volume
14
Issue
4
Pages
e1006114
ISSN
  • 1553-734X (ISSN-L)
  • 1553-7358
OpenAlex
W2517431276
Cited by
73
References
46
Citations by year
2026: 5 · 2025: 5 · 2024: 4 · 2023: 10 · 2022: 8 · 2021: 6 · 2020: 6 · 2019: 7 · 2018: 19
FWCI
46.02
Type
journal article
Language
en
Open access
Locations
  • PLoS Computational Biology (publishedVersion, cc-by)
  • Europe PMC (PubMed Central) (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • arXiv (Cornell University) (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • Figshare (submittedVersion, cc-by)
  • arXiv (Cornell University)
Funders
Templeton World Charities Foundation (#TWCF 0067/AB41)
Topics
  • Social Sciences › Arts and Humanities › History and Philosophy of Science › Philosophy of Biology and Scientific Explanation
  • Physical Sciences › Physics and Astronomy › Statistical and Nonlinear Physics › Complex Systems and Dynamics
  • Physical Sciences › Computer Science › Computational Theory and Mathematics › Cellular Automata and Applications
Keywords
integrated information, black-boxing, emergence, reductionism, spatial and temporal scales, Boolean networks, local maxima, levels of organization
MeSH terms
Cell Cycle, Computer Simulation, Models, Biological, Schizosaccharomyces, Systems Theory, Computational Biology, Systems Biology

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.