Uncertainty About Sentience and the Benefit of the Doubt

Original text Luciano Carlos Cunha · Volume II, p. 42 of the PDF edition · English translation

13.1. Moral consideration and uncertainty about whether a being is sentient

It is sometimes argued that, with regard to nonhuman animals (or at least some of them), we are justified in denying them moral consideration because there is no proof that they are sentient.

Animal advocates usually respond to this objection by pointing out that there is a growing body of scientific evidence establishing sentience in a large proportion of nonhuman animals, to the point that this is now the default position among scientists. This is expressed in declarations issued by the scientific community. For example, the Cambridge Declaration of July 7, 2012, attests that all mammals and birds, as well as invertebrates such as cephalopods, possess the neurological substrates that create the conditions for the emergence of consciousness⁵¹. In turn, the New York Declaration of April 19, 2024, attests that the empirical evidence indicates at least a realistic possibility of consciousness in all vertebrates and in many invertebrates (at a minimum, cephalopod mollusks, decapod crustaceans, and insects)⁵².

However, not even this entire body of scientific evidence will satisfy the skeptic. In this text we will first discuss the objection raised by skeptics. It will be argued that, as far as ethics is concerned, the burden of proof lies with the skeptic. That is, it is not the animals that possess a centralized nervous system who have to prove that they are sentient: it is the skeptic who has to prove that they are not (until then, we should give them the benefit of the doubt).

Next, we will discuss the position that moral consideration should be assigned in degrees: according to this position, the greater the degree of uncertainty about whether a being is sentient, the less it should count in our moral deliberations. Some arguments for rejecting this position will then be offered.

13.2. The skeptical objection

Consider the claim that we cannot be certain whether nonhuman animals are sentient for the simple reason that the only case in which anyone can be certain of this is their own sentience, since it is the only one they can experience directly. This is known as the Problem of Other Minds⁵³.

One possible response would be to agree that the only case in which each individual can prove sentience with absolute certainty is their own, but to maintain that we nevertheless have good reasons to believe that there are other individuals who are sentient. The first kind of reason is behavior. For example, if I step on a shard of glass, I pull my foot away because I am feeling pain. So if other individuals pull away every time they step on a shard of glass, that is a reason to suppose they are doing so because they too are feeling pain. The second kind of reason is physiology. If I know that I can have experiences because I have a nervous system with a brain, and I know that it is the processing that takes place in the brain that creates the conditions for consciousness to arise⁵⁴, then that is a reason to suppose that any other being with a nervous system that has a centralizing organ (a brain or ganglia) carrying out this kind of processing is also sentient. This is known as the Argument from Analogy.

However, this response will not satisfy the skeptic. He will say that it is a hasty generalization from a single case. Law (2003, p. 102-3) gives the following example: if I open a thousand cherries and find a pit in the middle of each one, I am still not justified in concluding that every cherry necessarily has a pit in the middle, and I would be even less justified in concluding such a thing if I had opened only one cherry. In the same way, the author concludes, generalizing from one's own case that other individuals who have brains and nervous systems and behave similarly are also sentient is like concluding that all cherries have pits from observing a single cherry. Thus, the skeptic concludes, we have no reason to think that there are other minds besides our own (since all the scientific evidence of sentience rests on the Argument from Analogy).

At this point one might ask who bears the burden of proof. The skeptic says that we have no reason to think there are other minds besides our own. However, the skeptic does not

offer any reason for thinking that there are none either. Who bears the burden of proof here? In epistemic terms, this is debatable. In ethical terms, however, the burden of proof certainly lies with the skeptic. This will be explained in more detail below.

13.3. The principle of the benefit of the doubt

If the fact that the only case in which we can prove there is sentience is our own were a good reason to deny moral consideration to nonhuman animals, it would equally be a reason to deny it to all humans, since we cannot verify through direct experience that they are sentient either.

There is a principle that grounds the view that, even if we are not certain that other beings with centralized nervous systems are sentient, we should nevertheless give them moral consideration: the Principle of the Benefit of the Doubt⁵⁵ (also known as the Precautionary Principle). The principle prescribes that, where there is reasonable doubt about whether or not a being is sentient, we should treat it on the assumption that it is sentient, because the probability of harm is then lower.

The reasoning behind this principle is as follows: if the being in question is sentient and we treat it as if it were not, the harm we cause will probably be greater than the harm that would occur if we treat it as if it were sentient and it actually is not. In short, the principle prescribes erring on the side of caution.

If the principle of the benefit of the doubt is reasonable, then pointing out that each of us can only be certain of our own sentience is not a good reason to deny moral consideration to anyone.

13.4. Degrees of consideration according to the degree of confidence about sentience?

There is a position holding that the degree of moral consideration we should give someone ought to depend on the degree of confidence we have about whether or not the being in question is sentient: if we are certain that it is, the individual counts as 1; if we are not, we multiply by less than

1 according to the degree of confidence we have about whether or not it is sentient⁵⁶.

On the basis of this position it could be argued that, since the degree of confidence in the sentience of, for example, humans or vertebrates in general is much greater than in the case of various kinds of invertebrates, we should give less consideration to the latter.

There are at least four responses to this objection, which will be addressed below.

13.4.1. Rejecting the degree-of-confidence criterion

The first response is to reject assigning differentiated degrees of moral consideration on the basis of our degree of confidence about whether or not the being in question is sentient. For example, one alternative would be to propose a minimum threshold of reasonable evidence and to maintain that everyone above that threshold would count as 1, without hierarchies, regardless of how close to or far from the threshold they are. Thus, proponents of a hierarchy based on degree of confidence would need to defend it, not simply postulate it, given that other possibilities exist.

Moreover, in many cases our low confidence regarding the sentience of certain kinds of animals stems from the scarcity of research on their sentience, or from biases on our part in assessing the data from the available research, and not from their clearly failing to meet the criteria for sentience⁵⁷. Thus, against the degree-of-confidence criterion, it could be argued that in a matter as serious as moral consideration, what we should do is establish safeguards against our lack of information and against our likely biased judgment, rather than use these limitations as a criterion.

13.4.2. Arguing that the degree of confidence is high

The second way of responding would be to argue that the existing evidence for the sentience of invertebrates⁵⁸ such as decapod crustaceans⁵⁹ and insects⁶⁰ is sufficiently

strong, and so it makes no sense to say that such animals should count as much less than 1 even if we accepted the degree-of-confidence criterion.

Of course, it may be that in the case of other animals, such as gastropods and bivalves, the available evidence is not as strong. But, again, this is not because they clearly fail to meet the criteria for sentience, but rather because it is not known exactly what the minimum number of neurons and of interactions among them is for sentience to exist⁶¹.

13.4.3. Pointing out that counting for less is not enough to justify exploitation

The third way of responding points out that, even if such animals counted as less than 1, it does not follow that exploiting them is justifiable or that we are justified in failing to help them. It is one thing to be able to establish that certain individuals should count for less. It is another to be able to establish that we are justified in exploiting, or in failing to help, the individuals who should count for less. An additional argument establishing this second conclusion would have to be put forward, since it does not necessarily follow from the first. For example, the probability that they are sentient may be high enough that we should protect them from being exploited and help them when they are in need.

13.4.4. The number of individuals can offset the smaller weight given to each

The fourth response is to observe that, even if each of these animals counted as much less than 1, the number of them who suffer and are killed is so enormous that we should still prioritize protecting them.

For example, suppose that each human counted 1,000 times more than each invertebrate. This would mean that, for us to invest the same amount of resources in helping invertebrates as we should invest in helping humans, there would have to be 1,000 times more invertebrates than humans in equally bad situations. Since in the real world the number of invertebrates who suffer and die is vastly greater than that (whether

by being exploited or in the wild), the balance would tip toward prioritizing protecting them even if it were postulated that each human counts many times more. However much smaller the weight assigned to each invertebrate, there would always be a number of invertebrate victims above which the balance would tip toward prioritizing them.

The only way to avoid this conclusion would be to set up a trump-like barrier that would forbid prioritizing invertebrates no matter how many more of them were harmed⁶². However, this move is ad hoc, because an additional argument would have to be put forward to ground this trump, since it does not follow from the claim that an individual's moral weight should be multiplied by our degree of confidence about whether it is sentient (on the contrary, in this case it aims to limit the implications of the degree-of-confidence criterion).

In this other text you can read an argument against the establishment of trumps or any other kinds of hierarchies of moral status.

13.5. Conclusion

In this text we have seen that there are good reasons to apply the principle of the benefit of the doubt and that, therefore, at least in ethical terms, the burden of proof lies with those who are skeptical about the sentience of beings that possess a centralized nervous system, whether with a brain or with ganglia. This also shows that, although a great deal of scientific evidence supports sentience in many kinds of animals, in ethical terms such proof is not necessary for us to owe them moral consideration: reasonable doubt is enough.

We also examined the position that the degree of moral consideration should depend on the degree of confidence we have about whether or not the being in question is sentient. We saw that it is possible to question the validity of the degree-of-confidence criterion, and we also saw that, even if we adopted this criterion, it would still not follow that we would be justified in exploiting the animals in whose sentience our degree of confidence is lower, nor that we would be justified in investing little in helping them. In fact, given their numbers, we should still prioritize helping them.