402. "Nothing is so certain as that I possess consciousness." In that case, why shouldn't I let the matter rest? This certainty is like a mighty force whose point of application does not move, and so no work is accomplished by it.
403. Remember: most people say one feels nothing under anaesthetic. But some say: It could be that one feels, and simply forgets it completely.
--Wittgenstein, Zettel (1929-1948)
I offer for LW's consideration the interesting 2008 paper "Inverse zombies, anesthesia awareness, and the hard problem of unconsciousness" (Mashour & LaRock; NCBI); the abstract:
Philosophical (p-) zombies are constructs that possess all of the behavioral features and responses of a sentient human being, yet are not conscious. P-zombies are intimately linked to the hard problem of consciousness and have been invoked as arguments against physicalist approaches. But what if we were to invert the characteristics of p-zombies? Such an inverse (i-) zombie would possess all of the behavioral features and responses of an insensate being, yet would nonetheless be conscious. While p-zombies are logically possible but naturally improbable, an approximation of i-zombies actually exists: individuals experiencing what is referred to as "anesthesia awareness." Patients under general anesthesia may be intubated (preventing speech), paralyzed (preventing movement), and narcotized (minimizing response to nociceptive stimuli). Thus, they appear--and typically are--unconscious. In 1-2 cases/1000, however, patients may be aware of intraoperative events, sometimes without any objective indices. Furthermore, a much higher percentage of patients (22% in a recent study) may have the subjective experience of dreaming during general anesthesia. P-zombies confront us with the hard problem of consciousness--how do we explain the presence of qualia? I-zombies present a more practical problem--how do we detect the presence of qualia? The current investigation compares p-zombies to i-zombies and explores the "hard problem" of unconsciousness with a focus on anesthesia awareness.
3. Inverse zombies
What would an inverse (i-) zombie look like? Since the p-zombie is a creature that behaves and responds as if it were conscious when in fact it is unconscious, we posit an i-zombie to be a creature that appears to be unconscious when in fact it is conscious. Any query of a p-zombie elicits a response to indicate consciousness; thus, any query of an i-zombie should thus elicit a response (or lack thereof) to indicate the absence of consciousness. Characteristics of the unconscious appearance of an i-zombie could be unresponsiveness to verbal commands, absence of spontaneous or evoked vocalization or speech, absence of spontaneous or evoked movement, and unresponsiveness to noxious stimulus. Like the p-zombie, the concept of the i-zombie entails no logical contradiction and hence can be considered both conceivable and possible. Unlike the p-zombie, however, i-zombies are naturally probable. We argue that a subset of patients experiencing awareness during general anesthesia, or ‘‘anesthesia awareness,” may fall into the category of i-zombie.
Having looked at some differences, we might also consider some similarities between p-zombies and i-zombies. It would seem that whatever solution we find for the problem of detecting consciousness in the case of i-zombies would be equally applicable to p-zombies in some important sense. What sense do we have in mind? In i-zombie cases, some type of consciousness detector could be used to confirm or disconfirm the hypothesis that anesthetized (or possibly even comatose) patients are conscious. In p-zombie cases, we could also use some type of consciousness detector to confirm or reject the same hypothesis with respect to infants, humans, animals, or aliens, which behave and function as if they are conscious. A consciousness detector of some sort would have to be able to distinguish between the presence and absence of consciousness in any possible creature and would therefore apply in detecting both p-zombies and i-zombies. Below we explore potential solutions to this consciousness detection problem (see Section 5).
4. Anesthesia awareness and anesthetic depth
Although the terms 'awareness' and 'explicit recall'are distinct and dissociable cognitive processes, in the clinical practice of anesthesiology 'anesthesia awareness' denotes both awareness and subsequent explicit recall of intraoperative events. Anesthesia awareness is a problem receiving increased attention by clinicians, patients, and the general public. A multi-center American study estimated incidence of awareness with explicit recall of approximately 0.13% (Sebel et al., 2004), a rate consistent with large European studies demonstrating awareness in 1-2/1000 cases (Sandin, Enlund, Samuelsson, & Lennmarken, 2000). A proportion of patients experiencing awareness may subsequently develop serious psychological sequelae, including post-traumatic stress disorder (Osterman, Hopper, Heran, Keane, & van der Kolk, 2001).
There are a number of subjective states that are associated with general anesthesia. In a recent study, dreaming has been reported in 22% of patients undergoing elective surgery (Leslie, Skrzypek, Paech, Kurowski, & Whybrow, 2007). Awareness itself can vary from the transient perception of conversations in the operating room to the sensation of being awake, paralyzed, and in pain (Sebel et al., 2004). The condition of anesthesia awareness is truly a clinical ‘‘problem of consciousness.” This can also occur in patients with neurologic injury leading to vegetative states or locked-in syndromes (Laureys, Perrin, & Bredart, 2007).
...These shortcomings led to the development of EEG techniques to assess anesthetic depth and detect consciousness. In the 1930s, it was demonstrated that the EEG was sensitive to the effects of anesthetics (Gibbs, Gibbs, & Lennox, 1937). There is not, however, a unique electrical signature that is common to all agents. Furthermore, the apparatus is bulky, labor intensive, and requires a dedicated observer in the operating room. Due to these limitations, processed EEG modules that often rely on Fourier transformation have been developed. Such 'awareness monitors' include the Bispectral Index, Narcotrend, Patient State Index, A-line, and others (Mashour, 2006). In general, these modules collect raw EEG and/or electromyographic data, subject them to Fourier transform, and then analyze parameters that are thought to best represent a state of hypnosis. The output is often a dimensionless number, usually on a scale of 100 (wide awake) to 0 (isoelectric EEG). One such monitor has been shown to reduce the incidence of awareness in a high-risk population (Myles, Leslie, McNeil, Forbes, & Chan, 2004), although the results of this study have recently come into question (Avidan et al., 2008).
...Such EEG-based monitors, although promising, also have limitations (Dahaba, 2005). Many of these modules are insensitive to well-known anesthetics such as nitrous oxide, ketamine, and xenon. These agents may be pharmacologically similar in their effect on the N-methyl-D-aspartate glutamate receptor. Conversely, EEG monitors can be sensitive to agents that do not suppress consciousness, such as B-adrenergic blockers or neuromuscular blockers. There are other ways by which such 'awareness monitors' can be confounded, such as individuals who have a congenitally low-voltage EEG, as well as patients who are hypothermic or hypoglycemic. Finally, such monitors are subject to artifact from other electrical equipment in the operating room.
The current limitations of assessing anesthetic depth entail that we have no completely reliable way to ensure the absence of consciousness in a patient undergoing anesthesia and thus there is a class of individuals who may appear completely unconscious and yet who are nonetheless conscious. Furthermore, despite advances in demonstrating intentionality in patients with persistent vegetative states (Owen et al., 2006), neuroimaging techniques are not practical or even possible for real-time intraoperative monitoring. In short, for all practical purposes, i-zombies are not simply possible or probable—they are known to exist.
5. Philosophical implications of i-zombies
Standard philosophical criticisms of behaviorism are built around conceptual considerations alone and sometimes appeal to intuitions that behaviorists would find question-begging. By contrast, the existence of an i-zombie implies a compelling, empirically based counterexample to behaviorism. An i-zombie is not only real, but has feelings without the possibility of behaviorally responding to stimuli. Therefore, feeling is not simply responding to stimuli.
...A plausible alternative to behaviorism is functionalism. Functionalism arose on the philosophical scene in response to the shortcomings of behaviorism and type-type identity theory. Functionalism holds that mental states are interdefined in terms of causal relations: the defining characteristic of any mental state P is the set of causal relations that P has with respect to inputs, internal mental processes, and behavioral outputs (Fodor, 2000; see also Churchland, 1996). Instead of characterizing the mind simply in behavorial terms, functionalists argue for the causal efficacy of mental states. For example, my belief that a tidal wave is about to form is caused in me by my perception of wave patterns characteristic of tidal waves; and in relation to my desire to preserve my life, the fear of a potential tidal wave will cause me to seek shelter. In contrast to type-type identity theory, functionalists do not hold that mental states can be identified exclusively with a single type of matter (e.g., the neural stuff that composes our brains), but instead maintain that mental states can be realized in any suitably organized system.
An explanatory advantage of functionalism is that it affirms the mental as the source of behavior causation by insisting that mind is defined in terms of function, or by what it does—an interdefined web of causal relations between inputs, inner processes and outputs. An explanatory weakness, however, is that by defining mind in terms of causal relations, functionalism is logically compatible with the absence of experience itself (Armstrong, 2000, p. 142; see also Chalmers, 1996; Churchland, 1996; LaRock, 2007)...In order to motivate functionalism within this practical context, we need to answer a basic question, such as: Where is consciousness caused in the brain?
Answering the 'where' question of consciousness in functionalist terms returns us to our discussion of anesthetic depth. In order to localize the neurophysiologic endpoints of anesthesia such as loss of consciousness, we should not use structural space but rather functional or phase space. Phase space, fractal geometry, and strange attractors are now being employed to characterize states of consciousness and anesthesia. In the late 1980s, Watt & Hameroff, (1988) demonstrated that phase space analysis of EEG reveals distinct attractors and dimensions for the waking state, anesthesia, and burst suppression. More recent work from van den Broek, van Rijn, van Egmond, Coenen, & Booij, (2006) confirms fractal dimensionality as a measure of anesthetic depth.
...Taken together, one answer to the 'where' question of consciousness and anesthesia is 'phase space'. This form of explanation is consistent with functionalism as it does not attach itself to a specific neural process or location, but rather considers the overall dynamic or 'functional' properties of the system. Furthermore, because it can be applied to EEG analysis of the anesthetized patient, it also holds promise in the detection of i-zombies in the clinical realm.
6. The hard problem of unconsciousness
It should be clear immediately that the hard problem of unconsciousness is fundamentally practical or clinical. The fact that there is no uniformly reliable method to identify or predict intraoperative awareness leaves us with a situation in which consciousness is truly a problem. Assuming 30,000,000 general anesthetics delivered every year in the U.S. alone, with an incidence of anesthesia awareness of approximately 0.15%, we are left with 45,000 patients each year who have not had the adequate suppression of qualia. If we include patients who dream during general anesthesia, the number of potential i-zombies increases dramatically.
This problem is not limited to the operating room: it is becoming clear that patients who carry a clinical diagnosis of persistent vegetative state are capable of 'responding' (as assessed by functional imaging) in a way that indicates both comprehension and conscious intentionality (Owen et al., 2006). This hard problem of unconsciousness—detecting the presence of qualia—is again relevant. Decisions of continued life support, as in the highly publicized case of Terry Schiavo, are often made on the assumption of an absence of qualia.
...The foregoing examples highlight the ethical dimension of the hard problem of unconsciousness. The demonstrated natural possibility of i-zombies has implications for our treatment of individuals presumed unconscious. How should clinicians behave in the operating room given the demonstrated incidence of 1–2 individuals/1000 that may still experience qualia during a surgery? Should we comport ourselves acknowledging that the patient has the capacity for suffering? Should we at least ensure that if qualia cannot be extinguished that suffering is minimized with adequate analgesia? Should we restrict our speech to that which is respectful to all patients, conscious or ‘‘unconscious”? These ethical implications seem to readily fall out of the possibility of i-zombies.
...A more controversial ethical question relates to life support for patients with a diagnosis of persistent vegetative state. Given recent data suggesting that these patients may somehow covertly experience undetected qualia, what are the implications? Do we need to further consider the possibility that patients with even more dire diagnoses such as coma or brain death could potentially be i-zombies? The ethical exploration of this question is beyond the scope of this essay and would have important implications for end-of-life decision processes in critical care medicine, as well as organ donation.
...Furthermore, there can still be brain activation during general anesthesia. For example, primary and feed-forward visual processing persists during general anesthesia, while higher order processing is interrupted (Imas, Ropella, Ward, et al., 2005a, 2005b). A study of auditory processing under propofol anesthesia has reached a similar conclusion (Plourde, Belin, Chartrand, et al., 2006). These findings further emphasize the need to assess which brain states are associated with qualia. Mere activation or arousal of the brain does not necessitate consciousness and may still be a feature of an unconscious being. Indeed, this question touches not simply on the detection but on the very definition of i-zombies.
The awareness rate is chilling. One wonders whether surgery rates would be significantly affected in everyone was aware of this; it's like that utilitarian puzzler 'how much would I have to pay you to torture you with amnesia afterwards?' but in real life.
- "The Zombie Preacher of Somerset"
- "Consciousness lost and found: Subjective experiences in an unresponsive state"
- Armstrong, D. (2000). The nature of mind. In B. Cooney (Ed.). The place of mind (pp. 136–144). Belmont, CA: Wadsworth.
- Avidan, M. S., Zhang, L., Burnside, B. A., Finkel, K. J., Searleman, A. C., Selvidge, J. A., et al (2008). Anesthesia awareness and the bispectral index. The New England Journal of Medicine, 358(11), 1097–1108
- Chalmers, D. (1996). The conscious mind in search of a fundamental theory. Oxford: Oxford University Press.
- Churchland, P. (1996). Matter and consciousness. Cambridge: MIT Press.
- Dahaba, A. A. (2005). Different conditions that could result in the bispectral index indicating an incorrect hypnotic state. Anesthesia and Analgesia, 101(3), 765–773
- Fodor, J.J.A. (2000). The Mind-Body Problem. In: J.C. II (Ed.), Problems in mind: Readings in contemporary philosophy of mind (pp. 118-129): Mountain View: Mayfield Publishing
- Gibbs, F. A., Gibbs, L. E., & Lennox, W. G. (1937). Effect on the electroencephalogram of certain drugs which influence nervous activity. Archives of Internal Medicine, 60, 154–166
- Imas, O. A., Ropella, K. M., Ward, B. D., et al (2005a). Volatile anesthetics enhance flash-induced gamma oscillations in rat visual cortex. Anesthesiology,
- Imas, O. A., Ropella, K. M., Ward, B. D., et al (2005b). Volatile anesthetics disrupt frontal-posterior recurrent information transfer at gamma frequencies in rat. Neuroscience Letters, 387(3), 145–150
- LaRock, E. (2007). Disambiguation, binding, and the unity of visual consciousness. Theory and Psychology, 17, 747–777.
- Laureys, S., Perrin, F., & Bredart, S. (2007). Self-consciousness in non-communicative patients. Consciousness and Cognition, 16(3), 722–741. discussion 742–725
- Leslie, K., Skrzypek, H., Paech, M. J., Kurowski, I., & Whybrow, T. (2007). Dreaming during anesthesia and anesthetic depth in elective surgery patients: A prospective cohort study. Anesthesiology, 106(1), 33–42.
- Mashour, G. A. (2006). Monitoring consciousness: EEG-based measures of anesthetic depth. Seminars in Anesthesia, Perioperative Medicine and Pain, 25, 205–210
- Myles, P. S., Leslie, K., McNeil, J., Forbes, A., & Chan, M. T. (2004). Bispectral index monitoring to prevent awareness during anaesthesia: The B-aware randomised controlled trial. Lancet, 363(9423), 1757–1763
- Osterman, J. E., Hopper, J., Heran, W. J., Keane, T. M., & van der Kolk, B. A. (2001). Awareness under anesthesia and the development of posttraumatic stress disorder. General Hospital Psychiatry, 23(4), 198–204
- Owen, A. M., Coleman, M. R., Boly, M., Davis, M. H., Laureys, S., & Pickard, J. D. (2006). Detecting awareness in the vegetative state. Science, 313(5792), 1402.
- Plourde, G., Belin, P., Chartrand, D., et al (2006). Cortical processing of complex auditory stimuli during alterations of consciousness with the general anesthetic propofol. Anesthesiology, 104(3), 448–457
- Sandin, R. H., Enlund, G., Samuelsson, P., & Lennmarken, C. (2000). Awareness during anaesthesia: A prospective case study. Lancet, 355(9205), 707–711.
- Sebel, P. S., Bowdle, T. A., Ghoneim, M. M., Rampil, I. J., Padilla, R. E., Gan, T. J., et al (2004). The incidence of awareness during anesthesia: A multicenter United States study. Anesthesia and Analgesia, 99(3), 833–839
These are not at all uncontentious claims.
Maybe this whole article is a stealthy way to assert that p-zombies are a meaningful idea.
Meh. Being conscious for surgery is significantly less horrifying to me than dying of a burst appendix. I'd probably be pretty saturated with analgesics anyhow.
This isn't particularly related to p-zombies, though. An inverse p-zombie is a thing that responds exactly as if it weren't conscious, but is conscious. The authors talk about the problem of "detecting qualia" - but if you could, if you could follow every atom and had perfect decision-making, detect the consciousness, it wouldn't be an i-zombie. An i-zombie isn't a sleeping person who forgets that they dreamt, it's a conscious rock.
This is an i-zombie at the behavioral level, so it isn't entirely unrelated. You're right about the complete-physical-description level, which is the philosophers' favorite. The authors do point out that the poorly anesthetized are not i-zombies at the functional level, which also implies that they're not, at the complete physical description level.
What puzzles me is why they appear to favor functionalism for (e.g.) pain, over type-identity theory. None of their main points depend on it. And the supposed "advantage" of functionalism, that it affirms the mental as the source of behavior causation, applies equally to type-identity.
Is it less horrifying than not being conscious during the surgery?
Is not being conscious during surgery particularly horrifying?
I was, perhaps opaquely, pointing out Manfred's false dichotomy - the choice is not between surgery with faulty anesthesia and dying; the choice is between surgery with faulty anesthesia, dying, and a myriad of other options (such as the authors' own suggestion: intensified investigation into neural correlates of consciousness, so failed anesthesia can be detected and the surgery aborted or the anesthesia improved).
That's reasonable, and I'd agree there are some good third options (though if the pain is managed to less than getting my teeth drilled into under local anesthesia, I wouldn't really care). We're pretty much faced with my dichotomy now, though. So my statement was the sort of boring one that the current cost/benefit of going into surgery for anything worth even a few weeks of life remains very similar.
I'm not an anesthesiologist though, so I can't agree with that. Maybe there are alternatives. I can imagine there being workarounds or trade-offs which make more sense in light of these thoughts about i-zombies, such as greater use of local anesthetics (where it's very easy to test whether it works, as opposed to blanket general anesthetics).
The dual of this approximated i-zombie is just a sleepwalker, not a literal, atom-for-atom-identical p-zombie.
"This is what it’s like waking up during surgery: General anaesthetic is supposed to make surgery painless. But now there’s evidence that one person in 20 may be awake when doctors think they’re under", Robson:
The survey is https://academic.oup.com/bja/article/118/4/486/3574495 (Given the described wording and the remarkably blase acceptance claimed, I'm left wondering a little if the respondents really appreciated the scenario being described - being gutted like a fish and feeling every last bit of it, so to speak.)
"Patient perspectives on intraoperative awareness with explicit recall: report from a North American anaesthesia awareness registry", Kent et al 2015:
"Incidence of Connected Consciousness after Tracheal Intubation: A Prospective, International, Multicenter Cohort Study of the Isolated Forearm Technique", Sanders et al 2017:
"IamA Anaesthesia Awareness survivor! AMA!", by ohnozombees:
"Anesthetizing the Public Conscience: Lethal Injection and Animal Euthanasia", Alper 2008: Alper reviews curare etc and argues that US lethal injections, because of the use of paralytics and slow potassium poisons rather than quick effective standard veterinarian-style sodium pentothal injections, is manufactured anesthesia awareness:
Daniel Dennett turns out to discuss precisely this problem in the context of curare/analgesics/anesthetics/amnestics in Dennett 1978, "Why You Can't Make A Computer That Feels Pain".
He also discusses an interesting detail of pain, "reactive dissociation". In my pain taxonomy, I split the various kinds of pain disorders into useful/motivating/qualia; the only combination I was missing was a kind of pain which is experienced as painful and yet was not motivating/aversive/unpleasant. "reactive dissociation" turns out to be just that - if morphine is administered after pain starts happening, people apparently frequently will report that the pain is excruciatingly painful, and yet they don't mind it.
Dennett throws in this disturbing anecdote in footnote 27:
Why did I read this right before I was getting my wisdom teeth pulled?
When I re-read the paper before making this post, I was relieved that it had not occurred to me at all before undergoing general anesthesia for my wisdom teeth a while before!
I woke up while having my wisdom teeth pulled. I was not in pain, just the ordinary discomfort of all dental procedures.
10 month bump!
One of the fun things about LW is commenting on old articles.
Also upvoting old comments.
pg248-298 of Ketamine: Dreams and Realities, "Journeys into the Fright World"
I suggest a better proof-of-concept.
Lock someone in a box that has a monitor or some other method of talking to them, but no method of them talking to the rest of the world. They are presumably still conscious, but it would be impossible for the rest of the world to tell.
One way to do this would be to drop them in a black hole. Once they pass the event horizon, they cannot communicate with you, but you can still communicate with them, at least for a little while.
Another example is the universe as a whole. It has no input or output. How could it? Where would they lead? It still has conscious beings.
I thought the link would lead to this.
I mean the entire universe, not just the part we see. If we're in a simulation, the universe would include the computer the simulation is running on, along with everything else.
On an intermediate class of anesthetics: "Surgical Patients May Be Feeling Pain—and (Mostly) Forgetting It: Amnesic anesthetics are convenient and help patients make a faster recovery, but they don't necessarily prevent suffering during surgery", Kate Cole-Adams:
"Do Artificial Reinforcement-Learning Agents Matter Morally?", Tomasik 2014:
Particularly germane is the discussion of the disunified nature of cognition and experimental demonstrations of learning still happening while the hippocampus is knocked out by an anesthetic-like drug like "When Memory Fails, Intuition Reigns: Midazolam Enhances Implicit Inference in Humans".