Induction and Overreach in the Science of Consciousness
Abstract
Induction shapes consciousness science as researchers infer experience, generalizing neural signatures, and project across contexts. While deduction sets ideals of certainty, induction guides experimentation and decision-making. This article provides a synoptic analysis of induction’s role in consciousness science by surveying its philosophical background, from Aristotle and Hume to Popper, Salmon, and Peirce, and its formalization in Bayesian and predictive-processing frameworks. This manuscript then examines how induction operates across neuroimaging, comparative cognition, and clinical contexts, highlighting successes (pattern detection, anesthesia monitoring, covert awareness detection) and structural risks (correlation–causation slippage, overgeneralization, anthropomorphism, and high-stakes misclassification). Recent sufficiency-oriented syntheses (e.g., Coppola, 2025; Mashour et al., 2020; Hohwy et al., 2020) exemplify both the power and the perils of inductive generalization, extending correlational findings into claims of ontological sufficiency. Building on recent critiques and adversarial comparisons of leading theories, this article argues for an explicitly deflationary program that keeps induction indispensable but bounded. The core tools are: (i) operational vocabularies distinguishing constructs (access, arousal, valuation, metacognition); (ii) transparent bridge principles linking observables to experiences; and (iii) prediction tables that register both positive predictions and falsifiers. This pragmatic framework does not solve Hume’s problem; it makes our inductive commitments tractable, testable, and ethically responsible. Consciousness science should remain inductive, just not inflationary. To demonstrate this framework’s tractability, the paper introduces examples of bridge principles and quantitative criteria for “bounded” inference, including Bayesian priors, model-comparison metrics, and causal graphs. Ethical and institutional dimensions are expanded to include AI, animal, and algorithmic induction. These additions link philosophical modesty with methodological implementation and moral accountability.
Related
- Distinct electrophysiological correlates of low- and high-level visual consciousness with minimized post-perceptual processing
- Focused Ultrasound Neuromodulation of the Central Lateral Thalamus in a Non-Human Primate Model of Disorders of Consciousness
- Consciousness Beyond the Brain: The Role of the Autonomic Nervous System in Non-ordinary States of Consciousness
- The consciousness-first commitment of integrated information theory.
- Analog Cognition and Consciousness
BibTeX
@misc{gunn2026induction,
title = {{Induction and Overreach in the Science of Consciousness}},
author = {Matthew Phillip Gunn and Christine C. Minton and Chienyem A. Nweke},
howpublished = {PsyArXiv},
year = {2026},
month = oct,
doi = {10.31234/osf.io/bf3ue_v2},
url = {https://osf.io/preprints/psyarxiv/bf3ue_v2/},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Induction and Overreach in the Science of Consciousness
AU - Matthew Phillip Gunn
AU - Christine C. Minton
AU - Chienyem A. Nweke
PY - 2026
DA - 2026/10/22
PB - PsyArXiv
DO - 10.31234/osf.io/bf3ue_v2
UR - https://osf.io/preprints/psyarxiv/bf3ue_v2/
AB - Induction shapes consciousness science as researchers infer experience, generalizing neural signatures, and project across contexts. While deduction sets ideals of certainty, induction guides experimentation and decision-making. This article provides a synoptic analysis of induction’s role in consciousness science by surveying its philosophical background, from Aristotle and Hume to Popper, Salmon, and Peirce, and its formalization in Bayesian and predictive-processing frameworks. This manuscript then examines how induction operates across neuroimaging, comparative cognition, and clinical contexts, highlighting successes (pattern detection, anesthesia monitoring, covert awareness detection) and structural risks (correlation–causation slippage, overgeneralization, anthropomorphism, and high-stakes misclassification). Recent sufficiency-oriented syntheses (e.g., Coppola, 2025; Mashour et al., 2020; Hohwy et al., 2020) exemplify both the power and the perils of inductive generalization, extending correlational findings into claims of ontological sufficiency. Building on recent critiques and adversarial comparisons of leading theories, this article argues for an explicitly deflationary program that keeps induction indispensable but bounded. The core tools are: (i) operational vocabularies distinguishing constructs (access, arousal, valuation, metacognition); (ii) transparent bridge principles linking observables to experiences; and (iii) prediction tables that register both positive predictions and falsifiers. This pragmatic framework does not solve Hume’s problem; it makes our inductive commitments tractable, testable, and ethically responsible. Consciousness science should remain inductive, just not inflationary. To demonstrate this framework’s tractability, the paper introduces examples of bridge principles and quantitative criteria for “bounded” inference, including Bayesian priors, model-comparison metrics, and causal graphs. Ethical and institutional dimensions are expanded to include AI, animal, and algorithmic induction. These additions link philosophical modesty with methodological implementation and moral accountability.
ER -