Embodied Mind: Neurophenomenology

Every experiment on consciousness depends on reports. A participant presses a button when she sees a stimulus, or says whether an image looked bright or dim. Without such reports, a researcher couldn't tell which brain activity went with which experience. Yet reports are often handled as a necessary evil: the minimum needed to sort trials, with anything else a participant might say about the experience set aside as noise.

There are reasons for the caution. Introspection has a poor record. Eric Schwitzgebel argues that naive introspection is unreliable even about current conscious experience, not only about its causes (Schwitzgebel 2008). People disagree, and are often confidently wrong, about basic features of their own experience, such as how clear their peripheral vision is or whether their thinking comes in words or images. If reports are unreliable, they look like a weak foundation for a science. The history of psychology seems to confirm the worry. Laboratory introspection around 1900 stalled partly because trained observers in different laboratories couldn't agree about what they found.

That leaves two responses. One is to use reports as little as possible, keeping them to simple judgments such as "seen" or "not seen." The other is to make reports better. The second response treats unreliability as a problem of method, not a fixed limit. Untrained observers may describe their experience poorly for the same reason untrained observers describe a tissue sample poorly: nobody taught them what to look for, or how. The question then becomes which methods improve reports, and how anyone could tell that they have.

Researchers have developed methods along these lines. One is an interview method developed by Claire Petitmengin, in which an interviewer helps a person describe a single, specific experience in fine detail (Petitmengin 2006). The interviewer steers attention away from general opinions about the experience and toward how it actually unfolded, moment by moment. Methods of this kind, now often grouped under the name "micro-phenomenology," are used in current research on attention, perception, and clinical experience.

For a researcher, the question matters because it's already answered, implicitly, in every study design. Each design decides what participants may report, how they're prepared, and whether their reports shape the analysis or only label the data. Asking those questions gives you a disciplined way to use experience as data, and a way to read any consciousness study for how much weight its reports can bear.

References

• Petitmengin, Claire. 2006. "Describing One's Subjective Experience in the Second Person: An Interview Method for the Science of Consciousness." Phenomenology and the Cognitive Sciences 5 (3–4): 229–69.
  • Free: Author's PDF
  • Publisher: doi:10.1007/s11097-006-9022-2
• Schwitzgebel, Eric. 2008. "The Unreliability of Naive Introspection." Philosophical Review 117 (2): 245–73.
  • Free: Author's PDF
  • Publisher: Duke University Press

Introduction

In 1995 David Chalmers argued that explaining the brain's functions would leave a further problem untouched: why those functions are accompanied by experience at all. A year later, Francisco Varela answered, not with a theory, but with a method.

This reading explains Varela's diagnosis of the hard problem, his working hypothesis of mutual constraints, and the example he developed from Husserl's analysis of time-consciousness. Citations to Chalmers's article give its section number, and citations to Husserl's lectures give section numbers (§).

The Diagnosis

Chalmers named the hard problem: Chalmers's name for the problem of explaining why and how physical processes are accompanied by experience at all (Chalmers 1995, sec. 3). He contrasted it with problems of explaining functions such as discrimination and report, which are hard in practice but not in principle.

Varela accepted that the problem was real (Varela 1996). What he rejected was the assumption that it could be solved by theory alone, by finding the right extra ingredient in physics or the right philosophical argument. On his diagnosis, the gap persists because science lacks a method for examining experience itself. Cognitive science studies the brain rigorously and treats experience casually, relying on untrained reports. The remedy he proposed was methodological: take experience as seriously as the brain, and study both with discipline.

He called the program neurophenomenology: Varela's research program in which disciplined first-person descriptions of experience and neuroscientific data constrain and refine each other. The shift here is from asking how the brain produces experience to asking how descriptions of experience and data about the brain can be brought to bear on each other.

Mutual Constraints

The program's core is mutual constraints: Varela's working hypothesis that first-person descriptions guide the analysis of neural data, and neural findings in turn refine the descriptions. It's a hypothesis, not a result. It bets that the two kinds of evidence can be made to fit, and that each can improve the other.

The constraint runs in both directions.

Neither side is reduced to the other. Varela didn't claim that experience is nothing but brain activity, or that neural data are only signs of experience. He claimed that each is incomplete without the other.

An Example: The Living Present

Varela developed the approach for time-consciousness (Varela 1999). His starting point was Husserl's analysis (Husserl [1928] 1991, §§8–13). For Husserl, the present isn't an instant. It's the living present: the present as a flowing three-part structure of primal impression, retention, and protention, not an instant. Retention, in Husserl, is the just-past held in the present experience and given as just past. It's part of the current experience, not a separate act of memory. Protention, in Husserl, is the open anticipation, within the present experience, of what is about to come.

Varela proposed that this description constrains the search for neural counterparts. If the present has duration and inner structure, researchers should look for brain processes that integrate activity over a span of time, not for an instantaneous state. He suggested that transient, large-scale synchronization of neural activity, lasting a fraction of a second to a few seconds, might be such a process. In the other direction, the dynamics of such processes might sharpen the description, for example of how the present's span varies.

Varela presented this as a research proposal. It illustrates the working hypothesis rather than confirming it.

Conclusion

Varela accepted the hard problem but diagnosed it as a gap in method, not only in theory. Neurophenomenology answers with a working hypothesis of mutual constraints: first-person descriptions guide the analysis of neural data, and neural findings refine the descriptions. His treatment of the living present shows how a phenomenological description can tell neuroscience what to look for.

Key Terms

• Hard problem: Chalmers's name for the problem of explaining why and how physical processes are accompanied by experience at all.
• Neurophenomenology: Varela's research program in which disciplined first-person descriptions of experience and neuroscientific data constrain and refine each other.
• Mutual constraints: Varela's working hypothesis that first-person descriptions guide the analysis of neural data, and neural findings in turn refine the descriptions.
• Living present: The present as a flowing three-part structure of primal impression, retention, and protention, not an instant.
• Retention: In Husserl, the just-past held in the present experience and given as just past. It's part of the current experience, not a separate act of memory.
• Protention: In Husserl, the open anticipation, within the present experience, of what is about to come.

References

• Chalmers, David J. 1995. "Facing Up to the Problem of Consciousness." Journal of Consciousness Studies 2 (3): 200–219.
  • Free: Full text on consc.net
• Husserl, Edmund. (1928) 1991. On the Phenomenology of the Consciousness of Internal Time (1893–1917). Translated by John Barnett Brough. Collected Works 4. Dordrecht: Kluwer.
  • Free: None; often free through a university SpringerLink subscription.
  • Publisher: Springer
  • Free companion: SEP, "Temporal Consciousness" (Dainton), §4.2 · SEP, "Edmund Husserl" (Zahavi), §4
• Varela, Francisco J. 1996. "Neurophenomenology: A Methodological Remedy for the Hard Problem." Journal of Consciousness Studies 3 (4): 330–49.
  • Free: None.
  • Publisher: Ingenta Connect
• Varela, Francisco J. 1999. "The Specious Present: A Neurophenomenology of Time Consciousness." In Naturalizing Phenomenology: Issues in Contemporary Phenomenology and Cognitive Science, edited by Jean Petitot, Francisco J. Varela, Bernard Pachoud, and Jean-Michel Roy, 266–314. Stanford, CA: Stanford University Press.
  • Free: None.
  • Publisher: doi:10.1515/9781503617421-012
  • Free companion: SEP, "Temporal Consciousness" (Dainton)

Introduction

A hypothesis of mutual constraints needs descriptions of experience precise enough to constrain anything. Varela held that such descriptions don't come naturally. They take method and practice.

This reading explains the first-person methods neurophenomenology draws on and the three-phase structure On Becoming Aware gives them. Citations to Husserl's Ideas I give section numbers (§).

First-Person Methods

A first-person method is a disciplined procedure by which a person attends to and describes their own experience, as opposed to untrained self-report. Varela drew on three kinds (Varela 1996).

1. Phenomenological reduction. From Husserl, Varela took the epoché: the suspension of the general thesis. It puts the world's existence out of action without denying or doubting it (Husserl [1913] 2014, §32). The general thesis is the everyday, unspoken belief that the world simply exists as it appears. Suspending it frees attention to notice how things appear. Varela treated the epoché less as a philosophical position than as a practical skill that can be learned and improved.
2. Contemplative training. Meditative traditions train attention to stay stable and to notice fine features of experience, such as the onset of a thought. Varela saw these traditions as a resource that Western science had largely ignored.
3. Second-person interviewing. A second-person method is a method in which a trained interviewer helps a person describe a specific experience, guiding their attention without suggesting content. The interviewer stands between the subject's own first-person view and the scientist's third-person view (Depraz, Varela, and Vermersch 2003).

The three differ in who does the work. In the first two, the person trains their own attention. In the third, a second person supplies the discipline the subject may lack.

The Structure of Becoming Aware

Natalie Depraz, Francisco Varela, and Pierre Vermersch set out a common structure beneath these methods (Depraz, Varela, and Vermersch 2003). They call it the basic cycle of becoming aware, and it has three phases.

Suspension, in Depraz, Varela, and Vermersch, is the first phase of becoming aware: interrupting one's habitual thoughts and judgments about an experience. It's the epoché made practical. You stop assuming you already know what the experience is like.

Redirection, in Depraz, Varela, and Vermersch, is the second phase of becoming aware: turning attention from the content of experience, what is experienced, to the act of experiencing, how it's experienced. You stop asking what you're thinking about and start noticing how the thinking unfolds.

Letting-come, in Depraz, Varela, and Vermersch, is the third phase of becoming aware: a receptive waiting in which one lets what isn't yet noticed come into awareness, rather than searching for it. The shift here is from active looking to receptive waiting. Searching tends to find what it expects. Waiting lets something new appear.

The table sets out the three phases.

The authors add that awareness gained this way must then be expressed in words and validated with others. That's where second-person and intersubjective checks enter.

Why Method Matters for Mutual Constraints

The working hypothesis needs descriptions that are stable across trials and specific enough to divide the data. Untrained reports rarely meet that standard. The methods in this reading are how neurophenomenology tries to meet it. Whether they succeed is a live question, and critics have pressed it hard.

Conclusion

First-person methods are disciplined procedures for describing one's own experience: the epoché as a practical skill, contemplative training, and second-person interviewing. On Becoming Aware finds a common three-phase structure in them, suspension, redirection, and letting-come, followed by expression and validation. These methods are meant to supply the precise descriptions that mutual constraints require.

Key Terms

• First-person method: A disciplined procedure by which a person attends to and describes their own experience, as opposed to untrained self-report.
• Epoché: The suspension of the general thesis. It puts the world's existence out of action without denying or doubting it.
• Second-person method: A method in which a trained interviewer helps a person describe a specific experience, guiding their attention without suggesting content.
• Suspension: In Depraz, Varela, and Vermersch, the first phase of becoming aware: interrupting one's habitual thoughts and judgments about an experience.
• Redirection: In Depraz, Varela, and Vermersch, the second phase of becoming aware: turning attention from the content of experience, what is experienced, to the act of experiencing, how it's experienced.
• Letting-come: In Depraz, Varela, and Vermersch, the third phase of becoming aware: a receptive waiting in which one lets what isn't yet noticed come into awareness, rather than searching for it.

References

• Depraz, Natalie, Francisco J. Varela, and Pierre Vermersch. 2003. On Becoming Aware: A Pragmatics of Experiencing. Advances in Consciousness Research 43. Amsterdam: John Benjamins.
  • Free: None.
  • Publisher: John Benjamins
• Husserl, Edmund. (1913) 2014. Ideas for a Pure Phenomenology and Phenomenological Philosophy. First Book: General Introduction to Pure Phenomenology. Translated by Daniel O. Dahlstrom. Indianapolis: Hackett.
  • Free: None. Cite by section (§), which is stable across translations.
  • Publisher: Hackett · alternative translation: Kersten (Springer)
  • Free companion: SEP, "Edmund Husserl" (Zahavi), §3
• Varela, Francisco J. 1996. "Neurophenomenology: A Methodological Remedy for the Hard Problem." Journal of Consciousness Studies 3 (4): 330–49.
  • Free: None.
  • Publisher: Ingenta Connect

Introduction

Neurophenomenology began as a proposal. Its best-known test came in 2002, when Antoine Lutz and colleagues used participants' trained descriptions of their own readiness to organize an analysis of brain activity.

This reading summarizes that study and Lutz and Thompson's account of the research program it launched.

The Problem: Variability Treated as Noise

Present the same stimulus to the same person many times, and the brain's response differs from trial to trial. Trial-to-trial variability is the difference in neural responses across repeated presentations of the same stimulus, which standard analyses average away as noise.

Averaging makes sense if the variability is random. But some of it may reflect the person's state just before the stimulus: how attentive, prepared, or distracted they were. If so, averaging throws away information about experience. The question is how to recover it. Brain data alone can't say which differences matter to the experience. Lutz and colleagues proposed that the participants could.

The Study

The study used a depth-perception task (Lutz et al. 2002). Participants looked at a random-dot pattern with no depth, which then changed into an autostereogram, an image in which a three-dimensional shape emerges when viewed the right way. They pressed a button as soon as they saw the shape. Electroencephalography (EEG) recorded their brain activity throughout.

Before the main recordings, participants did practice trials and described their experience of each one in open terms. From these descriptions, stable categories emerged. A phenomenological invariant, in Lutz and Thompson's usage, is a stable structural feature of experience that recurs across trials and that trained participants can reliably describe. Here the invariants concerned readiness: how prepared the participant felt as the shape appeared.

Three categories emerged:

1. Steady readiness. The participant felt prepared and attentive when the shape appeared.
2. Fragmented readiness. The participant felt prepared, but attention wavered or was interrupted.
3. Unreadiness. The participant felt unprepared, for example surprised or distracted.

In the main recordings, participants briefly described each trial, and each trial was assigned to a category. A phenomenological cluster, in Lutz and colleagues' study, is a group of trials sorted together by the participant's report of the same type of experience, such as steady readiness, before the neural data are analyzed.

What the Brain Data Showed

The researchers then analyzed the EEG separately for each cluster. Their main measure was phase synchrony: the degree to which the oscillations recorded at different sites keep a consistent timing relation, used as a measure of large-scale neural integration.

The clusters showed different patterns. In steady readiness, synchrony in the fast gamma frequency range built up over frontal sites before the shape appeared, and reaction times were faster. In unreadiness, that early pattern was absent, and reaction times were slower. Fragmented readiness fell between. The variability that averaging would have erased turned out to track the participants' descriptions of their readiness.

The Program

Lutz and Thompson generalized the study into a research program (Lutz and Thompson 2003). Its steps can be set out in sequence.

1. Train participants in first-person methods, so their descriptions are stable and precise.
2. Use their descriptions to identify phenomenological invariants.
3. Use the invariants to organize the neural analysis, especially of variability that would otherwise be treated as noise.
4. Use the neural results to test and refine the descriptions.

They present the study as a first step that realizes the working hypothesis of mutual constraints in one direction more fully than the other: the reports clearly guided the analysis, while the refinement of reports by data remained largely a goal.

Conclusion

Lutz and colleagues used trained descriptions of readiness to sort trials into phenomenological clusters before analyzing the EEG. The clusters showed distinct patterns of frontal phase synchrony and different reaction times, so variability usually treated as noise tracked experience. Lutz and Thompson turned the study into a program for joining first-person methods and neuroscience.

Key Terms

• Trial-to-trial variability: The difference in neural responses across repeated presentations of the same stimulus, which standard analyses average away as noise.
• Phenomenological invariant: In Lutz and Thompson's usage, a stable structural feature of experience that recurs across trials and that trained participants can reliably describe.
• Phenomenological cluster: In Lutz and colleagues' study, a group of trials sorted together by the participant's report of the same type of experience, such as steady readiness, before the neural data are analyzed.
• Phase synchrony: The degree to which the oscillations recorded at different sites keep a consistent timing relation, used as a measure of large-scale neural integration.

References

• Lutz, Antoine, and Evan Thompson. 2003. "Neurophenomenology: Integrating Subjective Experience and Brain Dynamics in the Neuroscience of Consciousness." Journal of Consciousness Studies 10 (9–10): 31–52.
  • Free: Author-hosted PDF (Thompson's site)
• Lutz, Antoine, Jean-Philippe Lachaux, Jacques Martinerie, and Francisco J. Varela. 2002. "Guiding the Study of Brain Dynamics by Using First-Person Data: Synchrony Patterns Correlate with Ongoing Conscious States during a Simple Visual Task." Proceedings of the National Academy of Sciences 99 (3): 1586–91.
  • Free: PubMed Central, PMC122234 · PNAS
  • Publisher: doi:10.1073/pnas.032658199

Introduction

Neurophenomenology stands or falls with the quality of its first-person descriptions. That's where its main method has developed, and where its critics have aimed.

This reading explains micro-phenomenology as a current interview method, the main criticisms of neurophenomenology, and how its treatment of reports compares with other approaches. It ends with a table of core terms and a list of further reading.

Micro-Phenomenology

Claire Petitmengin developed an interview method for describing experience in fine detail (Petitmengin 2006). Work in this line is now usually called micro-phenomenology: an interview method, developed by Claire Petitmengin and colleagues, that helps a person describe the fine-grained, often pre-reflective structure of a single, specific experience.

The method has several features.

1. A single experience. The interviewer asks about one specific occurrence, not about experiences of that kind in general. General descriptions tend to report beliefs about experience.
2. Evocation. The interviewee is helped to return to the experience in memory, re-living it closely enough to describe it.
3. From what to how. Questions steer attention from the content of the experience to how it unfolded: what came first, what was felt where.
4. No suggestion. The interviewer uses open questions and the interviewee's own words, so as not to supply the content being reported.

The method aims to reach aspects of experience that are lived but not noticed, which untrained reports miss.

Criticisms

Neurophenomenology faces three main criticisms.

Introspection is unreliable. Eric Schwitzgebel argues that naive introspection, meaning, in Schwitzgebel's usage, untrained introspective judgment about one's own current conscious experience, is unreliable even about basic features of experience (Schwitzgebel 2008). Neurophenomenologists reply that their methods aren't naive: training and interviewing are meant to correct exactly these errors. Whether training produces accuracy, or only confidence, is disputed.

Training may shape the reports. A participant trained to notice certain features may learn to report them, whether or not they were present. An interviewer, however careful, may steer. Defenders reply that open questions, the participant's own words, and checks against behavioral and neural data limit the risk. Critics answer that the risk can't be removed, only managed.

Results may not scale. Studies have used small numbers of participants, simple tasks, and labor-intensive training and interviews. It's unclear how far the approach extends to large samples and complex experiences.

None of these criticisms has settled the matter. They set the standards a neurophenomenological study must meet.

Comparing Methods

Core Terms at a Glance

Conclusion

Micro-phenomenology refines second-person interviewing to reach unnoticed features of single experiences. Neurophenomenology faces criticisms about the reliability of introspection, the shaping of reports by training, and scale, which remain open. Compared with third-person and introspective approaches, it gives reports the most work to do, and so it must meet the highest standard for them.

Key Terms

• Micro-phenomenology: An interview method, developed by Claire Petitmengin and colleagues, that helps a person describe the fine-grained, often pre-reflective structure of a single, specific experience.
• Naive introspection: In Schwitzgebel's usage, untrained introspective judgment about one's own current conscious experience.
• Neurophenomenology: Varela's research program in which disciplined first-person descriptions of experience and neuroscientific data constrain and refine each other.
• Mutual constraints: Varela's working hypothesis that first-person descriptions guide the analysis of neural data, and neural findings in turn refine the descriptions.
• First-person method: A disciplined procedure by which a person attends to and describes their own experience, as opposed to untrained self-report.
• Second-person method: A method in which a trained interviewer helps a person describe a specific experience, guiding their attention without suggesting content.

References

• Petitmengin, Claire. 2006. "Describing One's Subjective Experience in the Second Person: An Interview Method for the Science of Consciousness." Phenomenology and the Cognitive Sciences 5 (3–4): 229–69.
  • Free: Author's PDF
  • Publisher: doi:10.1007/s11097-006-9022-2
• Schwitzgebel, Eric. 2008. "The Unreliability of Naive Introspection." Philosophical Review 117 (2): 245–73.
  • Free: Author's PDF
  • Publisher: Duke University Press

Introduction

A neurophenomenology study has two kinds of evidence, and its value depends on how they're joined. Reading one well means asking where the first-person reports did work that the neural data couldn't have done alone. This reading walks through the best-known example, the 2002 readiness study by Antoine Lutz and colleagues, in four steps. It flags a common pitfall and ends with an annotated study map you can reuse for other studies.

Locating the Study

The study is Lutz, Lachaux, Martinerie, and Varela's 2002 article in the Proceedings of the National Academy of Sciences (Lutz et al. 2002). It's free on PubMed Central, so you can check each step against the article's methods and results sections.

Here is a summary of the study to start from.

• Aim. To test whether participants' descriptions of their own mental state could explain the trial-to-trial variability in brain responses that standard analyses average away.
• Task. Participants looked at a random-dot pattern with no depth, which changed into an autostereogram, an image in which a three-dimensional shape emerges. They pressed a button when they saw the shape. EEG recorded their brain activity.
• Findings. Trials sorted by the participants' reports of readiness showed different patterns of frontal gamma-band phase synchrony before the shape appeared, and different reaction times. Variability that looked like noise tracked the participants' descriptions.

Walking Through the Study

Step 1: Identify the first-person method and how reports were collected

Start with the reports, because everything else depends on them. Ask three questions. Were participants trained? How were reports elicited? When were they collected?

In this study, participants first did practice trials and described each one in their own words. The descriptions weren't sorted into categories the researchers had chosen in advance. Stable categories of readiness emerged from the descriptions: steady readiness, fragmented readiness, and unreadiness. In the main recordings, participants gave a brief report after each trial.

Notice what this design does. The categories come from the participants' experience, not from the researchers' hypotheses about it. That's the first-person contribution. A study that asked participants to rate their readiness on a scale the researchers designed would be doing something weaker.

Step 2: Identify how the reports structured the neural analysis

Next, ask what the reports did to the data. In a standard study, reports label trials after the analysis, or only mark whether a stimulus was seen.

Here, the reports came first. Each trial was assigned to a phenomenological cluster by the participant's report, and the EEG was then analyzed separately for each cluster. The shift is from reports as labels on data already analyzed to reports as the structure of the analysis. Without the reports, the researchers would have averaged all trials together, and the differences between clusters would have vanished into the mean.

Step 3: Identify what the neural findings added to the reports

Now ask what flowed back. The neural data showed that the categories weren't only verbal. Each cluster had a distinct pattern of phase synchrony: early frontal synchrony in steady readiness, its absence in unreadiness. Reaction times differed in the same order.

This adds two things. It gives independent evidence that the participants' distinctions were real, because the brain data and the behavior sorted the same way. And it describes readiness in a new vocabulary, as a pattern of large-scale integration built up before the stimulus. That description could, in principle, send researchers back to ask participants new questions: for example, whether fragmented readiness has sub-types that match variations in the neural pattern.

Step 4: Assess the design's strengths and limits

Finally, weigh the design.

Strengths. The categories emerged from participants' descriptions. The reports organized the analysis before it was run. The neural and behavioral results converged. And the study recovered information that standard averaging discards.

Limits. The study was small and the task simple, so it's unclear how far the approach scales. Training may have shaped what participants noticed and reported. And the constraint ran mainly one way: the reports guided the analysis, but the study didn't go on to refine the reports using the neural findings. Mutual constraint was realized in one direction and pointed toward in the other.

Key Considerations

Terms.

A common pitfall. The most common pitfall treats the reports as a simple label rather than a constraint on analysis.

• Labels come after; constraints come first. If reports only tag trials that were analyzed anyway, they add nothing that a button press couldn't. Here they determined how the data were divided before analysis.
• The categories weren't imposed. They emerged from open descriptions. Reading them as the researchers' categories misses the study's first-person contribution.
• Constraint is a matter of degree. Ask how much of the analysis would change if the reports had been different. In this study, a great deal would have.

Summary

The walkthrough moves from the first-person method, through the reports' role in structuring the analysis, to what the neural findings added, and to the design's strengths and limits. Here is an annotated study map to check your reading against, and to reuse for other studies:

References

• Lutz, Antoine, Jean-Philippe Lachaux, Jacques Martinerie, and Francisco J. Varela. 2002. "Guiding the Study of Brain Dynamics by Using First-Person Data: Synchrony Patterns Correlate with Ongoing Conscious States during a Simple Visual Task." Proceedings of the National Academy of Sciences 99 (3): 1586–91.
  • Free: PubMed Central, PMC122234 · PNAS
  • Publisher: doi:10.1073/pnas.032658199