The role of ethical reflection and dialogue in conceptualising animal welfare

This paper argues that ethical reflection and dialogue can assist in understanding what animal welfare is. Questions about animal welfare’s nature are thorny and contested. Responding to an essay by Donald Bruckner, the paper acknowledges that animal welfare is a type of normative value (namely prudential value) distinct from ethical value and that the methodology for determining prudential value is not simply reducible to ethical thought. However, it contends that connections between ethics and understanding wellbeing are closer than we might expect. The paper argues that: the broad conception of welfare we seek must answer to ethics; ethical responses can sometimes expose existing hidden or denied beliefs about prudential value/wellbeing; some judgements about wellbeing’s nature are internal to and thereby newly revealed in ethical judgements; understanding prudential value often requires sensitivity and responsiveness somewhat like that required in good ethical thinking. In these ways, and perhaps contrary to our assumptions, ethics (and philosophy) can help us conceptualise animal welfare.


Introduction
This paper examines connections between ethics and the nature of animal welfare. It argues that the process or method of understanding what animal welfare is, may legitimately and even valuably include ethical (or moral) reflection and dialogue. Questions about animal welfare's nature are thorny and contested. People variously believe welfare consists in subjective feelings, natural living, or biological functioning, or a combination of those things, or something else again (Beausoleil et al., 2018;Broom, 1991;Duncan, 1993;Fraser, 2008;Sandøe et al., 2003). Finding the appropriate methodology for conceptualising welfare has practical importance: differences about animal welfare's nature can influence views on how to conduct welfare science and how to treat animals. For example, we may judge a farming system with healthy but frustrated or deprived animals as good or bad depending on our conception of animal welfare.
I address the question of how to determine what animal welfare is by responding to an essay by philosopher Donald Bruckner (Bruckner, 2020), which discusses conceptual misunderstandings about welfare, the appropriate methodology for defining it, and the often neglected but essential role of philosophical thinking in analysis of welfare (see also Bruckner 2019). Bruckner argues that enquiries into the nature of animal welfare and the ethics of animal treatment are conceptually distinct enquiries. His argument is illuminating. Yet it and similar arguments may also suggest to some people that ethical modes of thought are irrelevant to understanding animal welfare. Many welfare scholars probably already believe that although wellbeing and ethics are connected-for example because animal welfare science is partly justified by socio-ethical concern for animals and/or because the nature of welfare informs ethical duties to them-working out what animal welfare is nonetheless can and should occur independently of (and perhaps always prior to) ethical reflection about what we owe animals.
I augment Bruckner's account (in ways Bruckner may agree or disagree with) with a further perspective. I argue that while it is important to clearly distinguish the two forms of inquiry, nevertheless we should appreciate the possibility that systematic efforts to determine what animal welfare is can and even sometimes should be informed by ethical modes of thought about animals. The paper therefore opposes the rather natural view that investigations into what constitutes animal welfare are totally separate from and/or always methodologically prior to ethical reflection on the treatment of animals.
The paper's main claims are: that the conception of welfare we seek must answer to ethics; that ethical responses can sometimes expose prudential value beliefs that were hidden or denied (e.g., due to disciplinary bias); that some judgements about wellbeing's nature are internal to and thereby newly revealed in ethical judgements; and (relatedly) that understanding prudential value often requires sensitivity and responsiveness similar in some ways to that required in good ethical thinking. Such epistemic qualities, I claim, may be found in disciplinary experts but also in lay people. The overall argument reveals another way animal welfare science needs philosophy (Haynes, 2008).

Distinguishing prudential value and ethics
In this and the following section, I outline key points in Bruckner's essay (Bruckner, 2020). His Animal Welfare Science, Varieties of Value and Philosophical Methodology argues strongly against a common tendency to conflate the normativity of animal welfare with moral norms. Bruckner's careful distinction between 'ethical value' and (what philosophers call) 'prudential value' sets the stage for my own arguments.
As Bruckner notes, many contemporary animal welfare scientists acknowledge that welfare is inherently evaluative, not merely empirical. Defining animal welfare is not a purely scientific task, for welfare does not refer merely to a true description of (say) a physical object, physiological state, bodily motion, behavioural pattern, or community's preferences. Nor can animal welfare be logically derived from descriptive facts alone. Instead, describing animal welfare involves saying not merely how things are but in an important sense how they 'ought' to be. Welfare science and investigation is inescapably normative. Understanding animal welfare's basic nature, Bruckner argues, is thus primarily a philosophical not scientific task, even when empirical judgements are involved.
Yet, he continues, we may readily conflate the normative element in wellbeing with moral normativity. That conflation problematically entails that determining animal welfare's nature is purely or largely a task for ethics. Bernard Rollin exhibits this confusion clearly: Questions of animal welfare are at least partly 'ought' questions, questions of ethical obligation. The concept of animal welfare is an ethical concept to which, once understood, science brings relevant data. When we ask about an animal's welfare…we are asking about what we owe the animal, and to what extent (Rollin 2015, p. 760).
But this is incorrect, since "[p]rudential goodness and moral goodness are distinct types of goodness" (Bruckner, 2020, p. 391). We can appreciate this conceptual difference, Bruckner continues, by noting that we can judge an animal's welfare in the absence of any question of ethical obligation. For example, we may intelligibly rate an animal's welfare as poor even when the animal cannot be helped.
Moreover, even if we knew everything about an animal's welfare, our moral duties to them raises a further question, requiring analysis of a different kind. In Bruckner's words, to "assume that humans should protect or promote the welfare of animals is a substantive moral assumption that requires defence and goes beyond questions strictly about welfare" (Bruckner, 2020, p. 391). Accordingly, it is an open question whether, say, industrial farming practices that cause poor welfare are justified or abhorrent. Being equipped with a full understanding of what animal welfare is is Theory Hedonism: pleasant and unpleasant mental states Preference satisfaction Natural living, species specific behaviours, telos Objective list theory, e.g., capabilities approach (Nussbaum, 2006): play, social interaction, control over environment, etc. Biological health and functioning Pluralism: some combination of the above Table 1 Competing theories of prudential (intrinsic) goods insufficient-we need additional ethical analysis to draw moral conclusions about animal treatment.
In contrast to ethical value, prudential 1 value relates to notions of wellbeing, quality-of-life, benefits and harms, and faring well or badly (Griffin, 1986). It concerns what is intrinsically, not just instrumentally, good or bad for an animal. Some measures, such as cortisol levels, may be mere indicators of welfare. Other factors (e.g., having space to run around in) might be considered instrumental goods that facilitate intrinsic goods. Table 1. lists some competing theories.

Philosophical methodology in animal welfare science
For Bruckner, determining the best prudential value theory involves primarily conceptual analysis rather than empirical analysis, even though the latter may inform the former. Conceptual analysis can involve consulting our 'intuitions' and adjusting them by considering examples and theoretical constructs. Bruckner's 'toy' example is determining the best 'conceptual theory' of a 'sandwich': we can conceptually analyse this object by considering our initial 'intuitions' and considered responses towards (for example) non-standard cases (e.g., ice-cream sandwiches), uncertain cases (e.g., lettuce replacing bread), and so on.
Using a back-and-forth process between theory and considered judgements-a process often called 'reflective equilibrium'-we can arrive at the best conception of a sandwich. For Bruckner, although this 'toy' example lacks moral significance, a similar methodology nevertheless applies to analysing animal welfare. Accordingly, to determine what animal welfare is we can test our intuitions or judgements about, say, whether an ill animal with no negative affective experiences has bad welfare, whether painless killing can harm animals, whether natural living is prudentially better than captivity-and so on.
Conceptual analysis, including of values, is often difficult. Various prudential value theories are contested. Progress in conceptual analysis, says Bruckner, is often facilitated by testing our ideas with others and using thought experiments to pump our intuitions (Bruckner, 2020, p. 392). A thought experiment could involve constructing examples of animals in contrasting situations to test our responses. Furthermore, the views of ordinary but informed people can benefit prudential value inquiry without being the final word (Bruckner, 2020, p. 393). 2 Bruckner contends that conceptual analysis of animal welfare is (contra Rollin and others) distinct from ethical analysis of what we owe animals: "questions of welfare are [not] necessarily questions of ethics" (Bruckner, 2020, p. 391). Now, a possibly natural conclusion to draw from this 3 claim is that ethical analysis and prudential analysis should be kept apart and that they have rather little in common-aside perhaps from involving broadly similar processes such as reflective equilibrium and testing our 'intuitions'. Thus, it may be felt that ethical modes of thought concerning animals and their welfare should enter the scene (at least in many or most cases) only after the nature of welfare has been determined by non-ethical normative analysis. Yet I suspect that ethics and prudential analysis have more in common than this view suggests, and that ethical styles and modes of thought can sometimes significantly inform our understanding of animal welfare. That is what I shall argue.

Taking stock
Two relevant, plausible claims emerge from Bruckner's discussion: (1) the concepts of prudential value and ethical value (e.g., duties to animals) are distinct; (2) the methodologies for determining prudential value and ethical value are distinct. (These claims do not imply that animal welfare and ethics are totally unconnected-again, for example, it is surely true that our ethical duties to animals are often affected by our understanding of animal welfare.) I agree that conceptually and methodologically conflating animal welfare and ethics is problematic. A crude conflation may, for example, draw our focus excessively to obligations to animals rather than allowing us to get to the heart of animal wellbeing. Furthermore, while ethical analysis may be informed by conceptual analysis of prudential value, it is assuredly underdetermined by it. Ethical arguments clearly introduce a range of other concepts and considerations beyond those related to prudential value.
However, I am still concerned that these important points may create the impression that animal welfare (and its analysis) and ethics are more conceptually and methodologically separate than they are or can be. In truth, they may well be connected in additional ways. Prudential value judgements and ethical responses sometimes shape, accompany, and illuminate each other. Further, ways of understanding prudential value and ethical styles of thinking can have a crucial element in common-which, as I shall later explain, amounts to a similar kind of sensibility or responsiveness-a fact that can easily be missed when, for example, our informative model is 'ordinary' conceptual analysis of sandwiches or other objects like chairs, clothes, and houses.

Prudential value/wellbeing must broadly answer to ethics
The first point to make is that, howsoever its details are cashed out, our conception of animal welfare as prudential value is shaped by ethics in the basic sense that it must answer to it. 4 Specifically, it must be so answerable by being relevant to some ways in which an animal can be wronged. If our conception of prudential value did not answer to ethics in that way, then it would not (except by accident) be relevant, for instance, to understanding important moral duties to animals or to animal welfare science's social raison d'être. Bruckner writes that "whatever conception of animal welfare we adopt in the end will have something to do with how [ordinary] people use the term 'animal welfare'" (Bruckner, 2020, p. 394, italics orginal). Yes-and one (usual) way that people use that term is a way that is relevant to the ethical treatment of animals. My point is not merely that welfare as prudential value happens to be so answerable, but that it must be.
There might be some kinds of prudential value that do not so answer to ethics. Suppose, for example, someone claims that certain states can be prudentially good for a car or an animal robot, such as having well-maintained engineering features. That person may claim that when those features function poorly, the car or robot has negative welfare, even if smashing the car or robot cannot wrong them (Coghlan et al., 2019). However, one may object that these functional features are not genuinely good for these objects, but only good for them, say, metaphorically. Nothing, we may think, is really good or bad for the car or robot. 5 But now consider a plant or a protozoon. Here, we might feel that it makes more sense to say that certain features or functionings are genuinely (not just metaphorically) prudentially good for these living things. Although some people do believe that such simple living things can be wronged (Taylor, 2011), we are not compelled to agree with them. Assuming we agree they cannot be wronged, we seem to have located a genuine kind of prudential welfare that does not answer (in the relevant way) to ethics. But that kind of wellbeing is not (usually) what we are seeking to discover when we try to understand animal welfare or do animal welfare science.
Bruckner claims that the concept of animal welfare is not "inherently moral" (Bruckner, 2020, p. 387), that "[j]udgments of prudence and morality occur on different axes," and that to "assume that humans should protect or promote the welfare of animals is a substantive moral assumption that requires defence and goes beyond questions strictly about welfare" (Bruckner, 2020, p. 391). There is a sense in which these claims are right. But we should avoid concluding that the right or best conception of prudential value is not broadly shaped by ethics.
Bruckner does say that "the protection and promotion of animal welfare is a leading moral issue, so there is a real importance to determining the best conception of animal welfare" (Bruckner, 2020, p. 392). My simple point here is that ethics determines what general conception of prudential value we should employ when investigating animal welfare's nature. In that basic and limited sense, at least, the relevant concept of prudential value is "inherently moral" after all. Unlike the concept of prudential value that applies to (let us imagine) cars, robots, plants, or protozoa, the concept of prudential value that we are interested in understanding is necessarily linked to a very broad notion of how animals morally should be treated and regarded. Perhaps many will agree with this-after all, it only says that ethics fixes the nature of wellbeing at a general level. Yet ethical modes of thought can, I shall suggest, also shed a more discriminating light on the ingredients of wellbeing for animals (and humans). Ethical responses can be usefully involved methodologically in analysis of prudential value. But this requires further argument, which occupies the remainder of the paper.

Turning to ethical responses to help understand animal welfare
Welfare scholars Daniel Weary and Jesse Robbins say the "border between animal welfare and other issues in animal ethics can be muddy" (Weary & Robbins, 2019, p. 38). Their essay suggests that ordinary people's concerns about animals in various situations could illuminate what animal welfare is, although it does not explicitly discuss the relevance here of ethical concerns, situations, dialogue, and responses. Bruckner relates a thought experiment from Robbins and others who, using animal vignettes, surveyed ordinary people about the relative welfare of (a) naturally living but unhealthy chimpanzees and (b) non-naturally living (captive) but healthy chimpanzees (Robbins et al., 2018, p. 3). Perhaps surprisingly, participants claimed that natural living was more important to chimpanzee wellbeing than were affective or hedonic states. This finding, Bruckner argues, can inform prudential value analysis by testing our 'intuitions'. Using similar processes, we could also test our 'intuitions' about, say, closely confined domestic animals that are healthy and 'hedonically happy' but unable to engage in natural behaviours.
However, I want to explicitly argue for the notion that we can also test and inform our judgements regarding welfare by engaging ethical modes of thought. This approach relies upon taking certain ethical responses as epistemically valuable in promoting moral understanding, as opposed to being 'mere gut reactions'. Although some philosophers (e.g., some foundationalists) would dismiss such ethical responses as irrelevant or even as impediments to moral understanding, many philosophers do take them seriously. This includes philosophers who advocate 'reflective equilibrium' approaches in ethics.
Some examples should illustrate what I have in mind. In the following two examples, we are to imagine being involved in a serious exploratory conversation about what animal welfare is with a thoughtful colleague. (I consider objections in the subsequent section.) Caged animal example Imagine that your colleague asserts that people, perhaps like those in Robbins et al.'s study, are simply misguided in thinking that unhealthy freeliving animals (say cows or pigs) are better off than healthy closely confined cows or pigs. Welfare, let us imagine your colleague declaring, is entirely reducible to hedonic states. Now, suppose you disagree. How might you reply to this colleague who favours a purely hedonic theory of wellbeing? One possible reply is to describe to her further examples of animals in various conditions and circumstances, from real-life or scientific studies or through meaningful thought experiments. You might also invoke certain opposing philosophical theories of wellbeing to support your position. Such moves are consistent with Bruckner's method of conceptual analysis of prudential value as detailed earlier.
However, a different (yet not incompatible) approach is to raise moral enquiries for your open-minded colleague to consider. One might, for example, ask her: What would you think of a person who keeps her dog (or pig etc.) confined in a small cage its whole life? Even if we hypothesise (in this thought experiment) that this dog has many more positive affective (hedonic) states than other dogs who socialize and play in the park, don't you think this treatment is still morally cruel to the caged dog? What sort of person (you continue) would treat their dog like that? Clearly, the invitation to the colleague being made here is to respond as many people surely wouldthat is, by thinking of this treatment, despite its plentiful hedonic satisfactions, as cruelly wronging the animal. 6 By posing an ethical thought experiment, then, we may see whether our open-minded colleague agrees and changes her view about welfare's nature, or whether she still prefers her purely hedonic theory of wellbeing.
Painless killing example Imagine next that you and your colleague (with the hedonic theory) are now disagreeing about whether death harms animals (McMahan, 2002). Some people believe death is not a welfare issue (Webster, 1994). Suppose your colleague insists that death does not harm because death abolishes all possible hedonic states (both positive and negative) that an animal might be in.
But suppose you think, on the contrary, that death itself can damage animal welfare (Yeates, 2010), precisely by depriving animals of valuable states they may have enjoyed but for being dead (Coghlan, 2018, p. 361). You tell her this, but she is not convinced. Now, you could, as before, seek to further test and explore your colleague's position in non-moral ways. For example, you may observe that death is usually regarded as a major harm for humans (unless they are suffering terribly), including for human infants who, presumably like many animals, have no concept of the future (Jensen, 2017, p. 620), because it deprives them of future goods. You may then invite your thoughtful colleague to consider whether she feels a need to reconcile such views. Again, this method could be part of a now familiar form of conceptual analysis of welfare.
But again, you may also seek to prompt ethical modes of response as another or an accompanying methodological approach, this time with the aim of illuminating the question that interests you both: whether death damages welfare. For example, if your colleague maintains, despite listening carefully to the above sorts of non-moral considerations, that death itself is never a harm for animals, you might invite them to consider philosopher Michael Lockwood's influential moral thought experiment. Lockwood imagined a company, 'Dispose-a-pup Ltd.', that offers painless lethal disposal and replacement of healthy older puppies for people who only want puppies not adult dogs (Lockwood, 1979, p. 168).
This treatment, presumably, will strike many people as wronging the disposed-of dogs even though (or because) they are kept perfectly healthy and happy and are killed quite painlessly when they approach adulthood. In responding to this thought experiment, it is at least conceivable that your colleague may come to appreciate 7 not only that this treatment wrongs the older puppies but also and at the same time, in opposition to her previous belief, that it wrongs them by depriving them of future goods (like positive hedonic states). She could thereby come to appreciate that death harms animals and their welfare after all.
Lockwood counted on a certain general moral reaction from his readers. It is true, however, that some people might not respond to our re-purposing of his thought experiment in the way I suggested. Perhaps, instead, the colleague's view in our example about death and welfare will remain unmoved. She may also offer further arguments in defence of her view. Nonetheless, it is easy enough to imagine that very many people will respond with shock, disgust, horror, anger, discomfort, etc. to the Dispose-a-pup example (or to similar examples). And it is also conceivable that in so responding, their views about prudential value may be altered or enlarged. 8 In other words, a person-another or ourselves-may elicit or have certain moral responses that (as in this imagined example) involve an appreciation of important features of prudential value that were hitherto denied, unrecognized, or just not held.
Accordingly, ethical responses could have a place in our considered judgements and analysis of what is intrinsically good and bad for animals. And exploring moral responses seems especially methodologically useful if it produces prudential value insights that may not otherwise have happened (or may have occurred much later). It may, it seems, be precisely through ethical modes of thought that we are awakened by a compelling sense of, say, an animal's loss and by some sense of the nature of that loss-as when, for example, an animal is confined, caged, or killed. We may then embrace positions about wellbeing we previously did not hold. Many other kinds of ethically mediated awakenings about the nature of welfare might be possible. If this argument is right, then not only can judgements about welfare's nature inform ethical responses, but ethical responses can also inform judgements about welfare's nature. Thus, these two modalities of thought are mutually informing.
I should stress that I am not suggesting the use of moral questions to elicit and test thoughts about prudential value should be used in every case of theorizing about prudential value.
For one thing, that approach would require the right circumstances. Nor is it intended to be definitive or the last word. Rather, a moral approach is best seen as an element of a broader, ongoing, and open-ended process of conceptual investigation. Nonetheless, ethical modes of thought seem able to sometimes play a complementary, valuable epistemological role in animal welfare analysis. 7 Again, I am assuming that she and we can take such moral responses as facilitating ethical understanding rather than being, say, mere reactions. 8 My general point here does not depend entirely on these examples. We could imagine someone responding to Lockwood's thought experiment without horror, outrage, shock, etc. They might call sympathetic responses to the puppies misguided or sentimental and might even use this ethical example to try to show that death does not harm animals. Clearly, the choice of using specific moral examples and indeed of adopting a moral approach at all to promote insight into prudential value depends on the individuals involved in the mutual exploration. Thanks to a reviewer for prompting this thought.
None of this implies that ethically mediated reflection on prudential value may never mislead us. Indeed, it may sometimes create confusion rather than clarity. But of course, so too can other kinds of reflections, responses, and reasoning processes that concern wellbeing. That is partly why careful analytical processes, such as reflective equilibrium, have been advocated by philosophers (including Bruckner). In the next section, I shall offer a more precise account of how ethical approaches might inform prudential value (welfare) analyses. This task involves replying to important objections to my argument.

Prudential judgements that are internal to ethical judgements
My argument so far may have failed to convince. For someone might object that we simply cannot know what wrongs an animal in the above sorts of ethical cases (cases that appear to turn partly on welfare) unless we already have a certain belief or insight about what good and bad welfare is. For example, when we respond by saying (in the above examples) that confining the hedonically happy dog is cruel or that killing the older puppy is wrong, our responses must surely be based on an existing belief that these kinds of deprivations harm welfare. Thus, it may be claimed that our sense that we have had a brand-new insight or that our views have genuinely changed is only apparent. Our objector may contend that prudential value analysis is prior to ethical thinking about animals which, at least in moral cases that involve reference to wellbeing, can only follow it.
As we shall see, I do not think this is always the case, but my initial reply is to agree that this sometimes occurs. That is, I acknowledge that in some cases certain ethical responses merely expose a prior belief or insight about prudential value, perhaps one we didn't clearly recognize or were denying. Nonetheless, revealing an 'intuition' or belief that we already had but that was hidden in this manner might still be useful to our efforts to understand animal wellbeing. After all, we know that biases frequently infect our thinking, and we can have biases about what welfare is.
Indeed, academic disciplines can generate and perpetuate biases and assumptions. Because philosophy largely ignored and denigrated animals for centuries, some philosophers have surely inherited superficial views about animal lives (Schmidt, 2011). Similarly, life sciences have harboured their own dubious assumptions about value and wellbeing (Rollin, 2006). In animal welfare science, some beliefs (e.g., that welfare is purely a scientific concept, that death is not a welfare issue, etc.) might be promulgated relatively uncritically. Then, ethical modes of thought that expose underlying beliefs at odds with superficial assumptions may be methodologically useful. Critically reflective ethical modes may be engaged in privately or in dialogue with open-minded others (Bruckner, 2020, p. 391). In addition, the views of the public (Bruckner, 2020, p. 394) about welfare (including views that spring from ethical reflection) may sometimes serve as a corrective to individual or group bias.
However, there is a more serious objection to my argument to consider. This objection says that the proposed approach of engaging ethical modes of thought to gain insights about prudential value is circular in a logically problematic way. According to this objection, I seem to be problematically claiming that the prudential value judgement P depends on a separate ethical judgement E which itself depends on the separate judgement P…and so on, viciously circularly. That would mean we could not gain novel insights or views about wellbeing by means of ethical responses in the way I suggested-a way that does not involve merely uncovering existing beliefs about prudential value but rather gaining entirely new ones.
The objection apparently puts my contention in jeopardy. Yet there is a way of understanding the argument that is not viciously circular. It involves noticing that prudential value judgements can sometimes be internal to or partly constitutive of ethical judgements. To explain this, take our caged animal example as described above. Suppose our response in that imagined scenario is that (all things being equal) an 'owner' would indeed cruelly wrong a dog by keeping it hedonically happy in a cage that deprives it of socialisation and play. To call it cruel in this manner is to employ a moral notion. But the conception of cruel in the example as I described it also contains a prudential value element (here, it is an evaluative notion about depriving the dog of certain 'natural' behaviours and activities). Thus, the prudential value judgement partly constitutes or is internal to the moral judgement that such behaviour is cruel. This understanding of the ethical-cum-prudential value judgement avoids the vicious circularity objection to my position.
What about the example of painless killing as discussed above? An objection here is that the ethical judgement (i.e., that killing the replaced animal for the sake of only having young puppies wrongs it) is saliently different since it does not involve a notion like 'cruelty'-a notion we might call a 'thick' or multidimensional notionwhich, as I just argued, can clearly be constituted by both ethical and prudential conceptual elements. Hence, the objection here may be that this sort of case lacks the relevant sort of co-constituted response or judgement and that therefore my claim cannot apply. To this objection, one could reply that such treatment is cruel after all, and that 'cruelty' has various meanings or applications. But another reply is that even if we do not want to call such killing cruel, the response that it would wrong the puppy to treat it like that can still be a response that contains both ethical and prudential value conceptual elements.
How so? Recall that our second scenario described a conversation with someone who, though holding that animals can have positive states that have prudential value, nonetheless denied that death could harm them. The example effectively involved inviting that person in the light of their position on death and welfare as we characterized it to form the judgement that it would wrong the animal to cause it to lose its future positive states by killing it. As Michael Lockwood noted, our response to the Dispose-a-pup thought experiment may include being "disgusted for the animal's sake" because of the way its short (if happy) life is exploited and "discarded" (Lockwood, 1979, p. 168).
Here again, it seems possible that a fresh prudential value judgement can be conceptually embodied in an ethical judgement (see also next section). The colleague's new judgement that the treatment wrongs the puppy is inseparable from her new judgement (which she opposed prior to responding to the moral thought experiment) that the puppy is harmed by being deprived of its future life. 9 It appears, then, that we may sometimes non-circularly receive prudential value insights through certain ethical responses even when certain multidimensional concepts such as 'cruelty' are not involved. 10 This helps explain how new understandings about welfare can arise for us when we have certain ethical thoughts or responses.

Relations between ethical and prudential value sensibilities
In this final section, I shall offer an account of what the sort of response in which a new prudential value judgement is internal to an ethical judgement might be like and how it could come about. This account involves the idea that ethical and prudential value judgements or responses can have further similarities that we may also overlook. Recall Bruckner's description of the methodological process of determining prudential welfare value as "conceptual analysis" (Bruckner, 2020, p. 391). We have considered why that description can be helpful. However, 'conceptual analysis' in the prudential value context is arguably a very different beast than 'conceptual analysis' concerning many other entities, such as ordinary items like sandwiches, chairs, clothes, etc. Certainly, conceptual analysis of some entities or things can be complex and/or difficult-consider, for instance, analysis of the concept of 'intelligence' in 'artificial intelligence'. But it is possible that conceptual analysis of certain values is qualitatively different from conceptual analysis of such things.
Value theory raises its own issues. In philosophy, the nature of value and of our knowledge of value is contested (for example, there is philosophical debate about whether prudential and ethical values are 'subjective' or 'objective' and whether grasping them requires certain affective responses or not). In any case, many philosophers believe that the epistemology of such value is qualitatively different from the process of gaining conceptual knowledge about many ordinary physical objects. It is not that we necessarily must abandon certain processes (such as reflective equilibrium) within conceptual analysis when it comes to these evaluative matters. But we might ask whether the nature of understanding ordinary descriptive facts or concepts is significantly different from that of understanding a normative or evaluative value notion like wellbeing.
Bruckner, we saw, speaks of 'intuitions' 11 as inputs into the conceptual analysis of both 'ordinary' things and prudential value. That may imply or suggest that there is no great difference between our judgements and responses in these respective cases. But now we may propose that a certain kind of judgement that is peculiar and especially important to understanding prudential value is one that involves a certain human responsiveness and sensitivity. Furthermore, we may propose that this sort of respon-9 Of course, an adverse ethical judgement in the pup disposal example might be based on other considerations unrelated to loss of future goods. But it could also be characterized by a response to things the pup will miss out on. 10 A person might nonetheless see this treatment as (say) callous (Lockwood, 1979, p. 168) or as robbing the pup. Such 'thicker' characterisations may carry both ethical and prudential value elements. 11 The term 'intuitions' is problematic for some philosophers, who may prefer other terms such as 'responses ', 'judgements', etc. siveness and sensitivity to prudential values on the one hand and to ethical values/ norms on the other may be akin or similar. So, something like an ethical sensibility may be required for having (or might at least promote) insights about prudential value. 12 What sort of sensibility, responsiveness, or sensitivity do I mean? That question raises hard philosophical issues and has many answers. But let me gesture towards some possibilities that should at least begin to indicate that understanding prudential value and ethical value might involve distinctive modes of thought not found in 'ordinary' conceptual analysis. There are many possibilities-I shall discuss only some, and then only very briefly. My first aim in this exercise is to expose and to question an idea we might have that forming sound judgements about certain values, including ethical and prudential value, is essentially (in certain respects) the same thing or process as forming sound conceptual judgements about certain matters of fact. My second aim is to suggest that recognising a certain difference there provides a further way of seeing how understanding what welfare is, and allowing ourselves to engage in ethical modes of response, could be related.
One possibility concerning how we gain insight in ethics derives from the philosopher Iris Murdoch, who in her famous essay 'The Idea of Perfection' argued that true ethical understanding depends on a certain quality of attention to the other, an attention that includes love and justice (Murdoch, 2014, p. 33). In Murdoch's view, a kind of engaged, patient attention is required to fully see the ethical reality of another, or of a thing or a situation. That could include understanding the ethical reality of an animal and its life. Murdoch's view has its adherents, although it is more controversial than the more general idea that insight into matters of ethical and prudential value require a degree of a certain sensitivity and responsiveness (which is also controversial). That said, no specific theory of the nature of understanding and conceptual analysis regarding ethical and prudential value matters will go uncontested.
For a number of moral philosophers (including but not limited to those in care ethics and ecofeminist traditions) ethical insight about people and animals can occur through responses such as compassion, care, and empathy (e.g., Gruen 2015). For virtue ethicist Shannon Vallor, a fuller ethical understanding is impossible without a certain recognisable sensibility. Vallor writes that moral understanding entails a kind of knowledge extending well beyond a cognitive grasp of rules and principles to include emotional and social intelligence: keen awareness of the motivations, feelings, beliefs, and desires of others; a sensitivity to the morally salient features of particular situations (Vallor 2016, p. 26).
.Again, I do not want to argue for a particular understanding of ethical insight. The important points I want to suggest here are first, that ethical insight may require a certain sensibility that differs from the qualities needed in ordinary conceptual analysis, and second, that a similar sort of sensibility might well apply to the process of gaining insight into prudential value. To help us appreciate this possibility, let us suppose we are conversing with someone who clearly understands what animal mental states, 12 I thank a reviewer for providing suggestions on developing this point. natural behaviours, and physiological functioning are, but who (unlike our thoughtful colleague in the earlier examples) has trouble seeing that mental states like pain, the absence of natural activities, and/or disordered functioning could be genuinely bad for an animal in a way that answers to ethics.
The person we are imagining here comprehends certain descriptive facts, but cannot intelligibly relate those facts to the (ethically) relevant prudential values. Perhaps this person is comparatively deficient in qualities like attention, empathy, and sympathy. In any case, we may see that their lack of a certain sensibility undermines their ability to appreciate what animal welfare is. Their relative lack of sensitivity might extend to related ethical matters. For example, it might be apparent that they cannot clearly grasp how confining or provoking an animal could sometimes cruelly wrong them. Such a person may have as much difficulty in the case of those animals as some of us do in seeing that a car or simple robot (or a plant or protozoon) can be genuinely harmed or wronged.
We might now feel that a person with such a dulled sensitivity about things that matter, about this range of values, would make an unhelpful conversational partner in a serious dialogue about the nature of animal welfare-including in discussions about prudential value that engage with ethical modes of thought. Despite being able to understand perfectly well ordinary descriptive facts and concepts about animals, their lack of sensitivity is a serious barrier to understanding both prudential value and moral value. Although most people are not so radically lacking in this sort of sensibility, some people, we might think, do have more insightful sensibilities concerning morality and wellbeing than do others.
If we now return to the core idea in the previous section-the way in which some prudential value judgements are internal to ethical responses which do not depend on a prior judgement about prudential value-we can perhaps see how the similar sensibilities that facilitate prudential value and ethical understanding can go together. Let us imagine that the doubting but thoughtful colleague described earlier responds to the specific examples involving the caging or killing of a dog (or pig etc.) with the thought, 'What a terrible thing to do to an animal!' Internal to and inseparable from that new ethical response in our colleague may be the similarly new thought, "What a terrible loss for the animal!' This latter response is a new judgement (and potential insight) about prudential value or welfare.
I have phrased these responses just now to indicate that they (can) involve a certain sensitivity (perhaps forms of attention, empathy, compassion, or something akin) towards both salient ethical and prudential value features of particular cases. Thus, we can now suggest not only that ethical and prudential value judgements sometimes involve similar and important kinds of sensitivity, but also that sensitivity to ethical values and sensitivity to prudential values can sometimes occur in the very same response (as explained in the previous section). And such a sensibility does not only occur in disciplinary experts, but in lay people too (just as it may be lacking in either). That is one important reason for taking seriously views about welfare that belong to a range of people, not just scientists or disciplinary experts.
The forgoing gives us one reason to think that the conflation of ethical and prudential value questions by scholars (like Rollin) is, although problematic, nonetheless to some degree understandable. Although we should recognize the important differ-ences between the normative activities of understanding ethical values and understanding prudential value, we should also be open to the important connections and overlaps. I have argued that this recognition can strengthen and deepen the analytic process of conceptualising animal welfare. Neither philosophers nor scientists nor other experts, then, should rule ethical reflection and dialogue out of court when they set forth methodological processes for determining the nature of animal wellbeing, of what is intrinsically and genuinely good and bad for them. On the contrary, we have reason for recognising a valuable methodological place (however modest) for ethics in gaining insight into what animal welfare is.

Conclusions
Expanding on an essay by Donald Bruckner, this paper sought to contribute to debate about the nature of animal welfare and the appropriate methodology for understanding or conceptualising it. I endorsed the view that animal welfare is a type of normative and prudential value irreducible to ethical value, and I agreed that the methodology for determining prudential value and ethical norms or responsibilities to animals display important differences.
However, I argued against the possibly natural conclusion (or existing belief) that ethics cannot inform our understanding of what animal welfare is. I even suggested that ethics might sometimes have a valuable role to play in that analytic process. After claiming that the conception of prudential value we seek in these contexts is one that must broadly answer to ethics, I went on to argue that ethical responses can sometimes reveal prudential value beliefs that were hidden or denied, perhaps due to bias or unexamined disciplinary assumptions. More importantly still, I argued that there are occasions in which new judgements about prudential value are internal to, and in that way arise with, ethical responses and insights.
I next suggested that we should not assume that the appropriate methodology in understanding prudential value is simply 'ordinary' conceptual analysis. Rather, we might be open to the idea that it is often a type of analysis that requires sensitivity, responsiveness, and insight of a kind that is similar in some ways to that required in good ethical thinking. Indeed, when prudential value judgements and ethical judgements are conceptually co-constituted and arise together in the way described above, the sensitivity towards prudential value is then an element or a constituent of the sensitive response to ethical value (e.g., duties to animals). This connection may be missed when we adopt 'ordinary' conceptual analysis as our informative model for prudential value (welfare) investigations.
Prudential value analysis that engages ethical styles or modes of thought can be done privately or in dialogue with others whose perspectives on prudential value are worth taking seriously because they have the right sorts of responsiveness and sensitivity to such values. This can include various scientists and philosophers but also lay people lacking disciplinary expertise. Finally, the paper's arguments show that animal welfare science needs philosophy to, for example, help illuminate and explore its fundamental concepts.