Does the Rose review change our perspective on fish pain?


In recent news, the popular and angling press have seized upon a review written by Professor James D. Rose of Wyoming University (and colleagues) regarding the ability for fish to feel pain.


Unfortunately, the way that the review has been represented is that of a definitive and conclusive work, but for anyone with access to the paper it is apparent that this is not the case.

In effect, the review reads as an opinion piece, with little to add to the current train of thought on the matter. In it, Rose appears to take pains to address the current available evidence, citing numerous experiments that have taken place, drawing upon research performed, and concluding that the available evidence is not convincing in providing ample evidence for pain in fishes.

However, there are problems with this. Most obvious amongst them is that Rose doesn’t really state what it is that would make a good case for evidence of pain in fishes. Early on, there are telling lines that seem to slip through, which may reveal a little more on his requirements. By page five, he writes; "pain is a subjective experience that cannot be directly observed.

In response to this, Dr. Lynne Sneddon (pictured) of Chester University, at most of whose work the review is aimed, comments; "One cannot directly prove what an animal feels, but equally Rose has ignored the fact that you cannot then disprove it and it would be unscientific to do so."

This is rather important, as by throwing this into the mix, Rose makes it rather clear early on that the argument is actually in danger of slipping into the realms of philosophy of mind, rather than remaining in that of science. It’s almost as though Rose has set a ‘safe retreat’ position into which he can fall back as the evidence starts to pile up.

To be crude, if we’re talking of pain as an entirely subjective experience that cannot be observed, then we have a dilemma. After all, when it comes to individual qualia and sensation, there are few subjects that have as little a study sample in the world of science. If I wanted to be childish, I could even state that the only pain experience in this world I can vouch for conclusively is my own. Everyone else in the world could be faking, and I’d be none the wiser. I have a study sample of one, and that doesn’t make my position strong.

Rose seems adamant on insisting for stronger and stronger arguments throughout the work, raising the bar on what can be submitted as evidence for pain in fishes. It’s even more than that, as Rose seems to set the bar so high that the only animals he seems happy to ascribe a pain perception to are humans and (possibly) some higher mammals. The criteria he states even leave animals such as birds as incapable of conscious pain perception.

Rose seeks to address five issues throughout the review, and these are:

Proper conduct of pain research with fishes, including matters of experimentally assessing pain with valid measures;

Technical and interpretational problems that undermine studies purporting to have demonstrated a capacity for pain awareness in fishes;

Evidence from a wide variety of experimental and field studies that were not necessarily designed to study pain but offer insights into the possibility of pain experience by fishes;

Claims for conscious awareness in fishes;

Costs to humans and fishes of invalid definitions and mistaken beliefs concerning fish pain and suffering.

Some of these he seems to apply more weight to than others, and the final point, the cost to humans, barely gets a mention at all. Here, it is alluded to that ideas of pain in fish can potentially be damaging to welfare issues, but doesn’t really lay down an explanation of how. There’s a little flip-flopping about, followed by an attack on the ‘benefit of the doubt’ position, implying that to give benefit of the doubt to the matter is costly to the organisms involved, though the only example offered is one that states that assuming fish have a capacity for pain perception hinders further scientific progress involving fish.

Rose quite explicitly mentions; "The benefit of the doubt dogma is not benign, nor does it best protect fish welfare." He mentions the risk of mandating welfare policies on a ‘but what if they can’ basis, and this in itself is rather telling. It certainly suggests to me that in matters of doubt pertaining to an organism feeling pain, Rose suggests that the default position should be that they can’t. To me this is troubling, and though some other authorities also feel uncomfortable with the benefit of the doubt position, it is one that is ethically compelling.

The first issue addressed, involving matters of experimentally assessing pain, is seemingly devoted to drawing a clear boundary between nociception and pain. Nociception effectively refers to the ability of the body to transmit harmful events (burning, trauma, etc) to the brain, and pain, here defined, takes place within the brain, which may or may not be based around events involving nociception.

In setting up a case to reinforce the separate nature of pain and nociception, Rose steps into some territory that to me seems a little baffling in its outcome.

He refers to the human model to make these points, and even here the ground is shaky. He cites; "Pain often, but not always accompanies nociception; pain sometimes occurs without nociception." This is all well and good, and there are certainly cases of human nociception activation that cause a response, as he goes on to demonstrate in decerebrate humans, those in permanent vegetative state, and so on. Pain occurring without nociception stimulation is also well known, and anyone familiar with the concept of phantom limbs, where an amputee complaining of pain in a no longer extant leg, can testify to this.

But Rose seems to make much of this point, almost hankering that the two can be mutually exclusive. But pain does indeed often accompany nociception, and perhaps enough to confidently state that the two are rather intimate, though the crux of Rose’s argument is that we should be wary of anything identifying nociception and confusing it for pain.

This is fine, but as I’ve already stated earlier, this is just raising that bar higher and higher, until eventually we hit the point where nothing short of the impossible task of demonstrating the actual qualia experienced by fishes will do.

For myself, a slightly vague line is where Rose says; "When an animal model is being used to investigate some aspect of pain, it is vital to know that the model system is valid for the purpose." This strikes me as rather close to circular reasoning. To take this mentality to an absurd and oversimplified conclusion, I feel it’s a little akin to me saying that we can only test to see if an animal feels for pain if we know that animal can feel pain, but that’s the very point being contested.

A considerable chunk of the review, which seems somewhat pejorative in tone throughout, is aimed at discrediting existing research on nociception in fishes, frequently aimed at the works of Dr. Lynne Sneddon, and she has much to say on the subject.

She takes a similar view to the one I have formed from reading the review, stating; "the same argument is used with no new perspectives: if fish do not have an identical brain to humans then they are not consciously aware and cannot 'feel' pain. Accepting this opinion means that no animals (except possibly primates) can experience pain, including dogs, cats, horses, birds, reptiles and amphibians." Dr. Sneddon accuses Rose of taking a "biased and highly anthropomorphic view of pain."

Professor Paul Flecknell of Newcastle University also mentions this point. "The main argument made in the review," he tells me, "is that fish lack the neuroanatomical structures currently considered to be required for conscious perception of emotions such as pain in humans."He adds; "This same argument could be made for all non-human animals, since they lack to a greater or lesser extent the structures present in people."

Asked to sum up the review, Prof. Flecknell tells me; “Rose doesn’t add much to the debate, and it remains just that – he also doesn’t stray into the other reasons why we might want to treat non-humans in ways that assume they could experience pain. I prefer to keep an open mind on the issue of pain and sentience, and when it might emerge as a property of complex nervous systems – especially as I don’t know how complex a nervous system needs to be to have these properties.”

Dr. Sneddon is also critical of the way she and her work is represented throughout, telling me that the authors “have included a number of incorrect facts in this review about my published studies to fuel their biased opinion.”

She also feels it is “important that readers are aware that this is just an opinion piece, as the views of this small number of authors are dangerous should the fish industries use this as a justification to neglect the welfare of these important vertebrate animals.”

Aside from attacking the methods used to gain positive outcomes of nociception research in fishes (and using some unusual leaps back and forth between elasmobranchs and teleosts, even though elasmobranch research in this area is precariously thin on the ground) which Dr. Sneddon is quick to note misrepresents her position and cherry picks information useful to Rose’s cause, Rose also dwells on his fourth issue of the review – consciousness in fishes.

This is an important point for Rose, and one where a little of a strawman argument is then set up. On page 26, he states that “There seems to be an assumption made by those advocating the belief that fishes are conscious that if this assumption were true, it would automatically be justifiable to assume that their consciousness would be human-like enough to conclude that fishes experienced human-like pain and suffering. There is no basis for this belief other than pure speculation.”

Well, no. This again is an opinion, and one that I suspect Sneddon and others would disagree with. At no point am I seeing any research that claims the potential fish experience of pain is like our own, only that they experience something pain-like and ‘more than just nociception.’ This is, to me, a convenient way of belittling the opposition, suggesting that they are implying considerably more than they are.

Regardless of this position, Rose earlier remarks; "the absence of a clear statement of the proposed nature of fish consciousness renders the construct conceptually amorphous." This is an interesting comment, because it again assumes the very point under discussion. The nature of the research that is taking place is in part to determine what that very nature of consciousness is. To say that it cannot be discussed because it hasn’t yet been clarified is rather uncharitable, I feel.

Rose then goes on to discuss differences between fish and human consciousness, and the problems therein. But this also seems to be missing something. Having laid down the criteria that fish consciousness is undefined, and then trying to compare what we don’t know about fish to what we do know about humans seems further effort to muddle to two fields together, as though they should be somehow symmetrical. Such anthropomorphism seems to present itself repeatedly during the review.

Dr. Sneddon does have a response to issues of consciousness in fishes, remarking that one of the key features in definitions is that of self-awareness – something troublesome to show in a lot of organisms.

Typical methods of defining self-recognition involve mirror tests, and observable ability for and organism to recognise itself. Fish notoriously tend to attack or retreat from their own reflection when presented with it, and given that the underwater environment lacks reflective surfaces, this is hardly surprising. Unlike terrestrial animals drinking from water, fish have had very little evolutionary opportunity to deal with or somehow comprehend a reflection of themselves.

However, Dr. Sneddon points to an obvious exception here, in "Cichlid fish can recognise their own odour distinct from others but also distinct from closely related kin. Therefore, this is evidence for self recognition and the ability to discriminate one’s own smell from others."

Rose concludes the review with some statements that are subjective to the authors, but aired as though authoritative. Rose talks of  “the necessity of using definitions and measures that validly distinguish between nociception, the unconscious sensory detection of injurious stimulus and conscious pain. Our examination of the research literature revealed that these requirements have not been met in research leading to claims for fish pain."

He goes on to say: "One of the most conspicuous shortcomings in discussions of scientific evidence for fish pain has been the selective consideration of evidence," the very same point that researchers of the subject are accusing this very review of. Dr. Sneddon mentions several key features of text where Rose has omitted research or results, or failed to adequately explain differences in method before using comparative cases.

To sum up Rose’s review in his own words, he says that "prolonged consequences of nociceptive stimulation and injury, especially conscious pain, are highly unlikely." He accepts that nociceptors are present in fishes (albeit in lower numbers than mammals) but concludes “Our consideration of the available evidence leads us to conclude that fishes for which the behavioural, physiological, and neurobiological evidence is available, are unlikely to have the capacity for the full range of nociception that can cause agonizing pain in conscious humans.”

Regarding consciousness, Rose states that “Our assessment of these claims leads us to conclude that neither their rationale nor their supporting evidence is compelling, much less neurologically feasible.”

So basically, for readers that have followed this article so far, the outcome is ‘based on what we have, we don’t know,’ which is pretty much where we were before the review was published. Rose doesn’t feel that there’s enough evidence for conscious pain (having picked the parts of research that support this view) but at the same time hasn’t performed any actual experiment to show that fish cannot feel pain. There really is nothing new added apart from a tinkering of the facts to fit an outcome.

The simple point is that this is one review amongst many, with a role to play, but having found nothing decisive. A broader analysis of the evidence may have produced something more compelling in either direction, but we don’t have a broad analysis, we have a rather closed one.

As a final thought, I offer some words from the concluding remarks of Dr. Sneddon’s response. “I am rather surprised that a journal such as Fish and Fisheries would publish a biased opinion piece such as this and since this personally attacks my research I would have thought it only fair that I be given the opportunity to review it before it was published. I could have pointed out many errors and omission of important references to provide a more balanced text. As scientists, many of us are funded by public money and it is important that we take an unbiased view presenting our results, and allowing the public to make up their own minds rather than trying to force through a personal agenda.”

The debate continues, it would appear.