Morality in the Real World: Evolving Thoughts

by Luke Muehlhauser on March 29, 2011 in Ethics,Podcast

Alonzo Fyfe and I are still researching the next episodes of Morality in the Real World.

In the meantime, we’ll share some of our evolving thoughts about morality and desirism.

 

Alonzo’s Thoughts

I thought I would tell you what Luke and I have been doing with respect to Morality in the Real World.

We said we are doing research. Well, here’s some of the specifics…

The next episode in the podcast concerns the pleasure theory of desire. We were writing the script to that episode and we came upon a question.

See, Tim Schroeder, whose writings we have been using as a foil, suggests that pleasure is a measure of the difference between perceived desire satisfaction and expected desire satisfaction.

This is why we are “pleased” when things turn out better than expected and “displeased” when they turn out worse than expected.

Project managers recognize this fact. They know that one of the most important tasks in managing a project is “managing expectations”. If you promise them a 50-hedon solution to their problem and deliver a 40-hedon solution, they are disappointed and review your work poorly. While, if you promise a 25-hedon solution and deliver a 30-hedon solution you have exceeded expectations.

This is also useful for an employee trying to please a boss.

I’m using “hedons” here as a measure of perceived value. This post here isn’t a formal addition to the Morality in the Real World podcast. It is an overview of where things are going.

Anyway, I went into Schroeder’s book to find the neurscience that supports this theory and . . . he didn’t have any.

So, Luke wrote to Schroeder to try to get an answer. He gave us some evidence, but it wasn’t as solid as we would have liked.

So, we suspended the podcast until we figured out what the neurscience of pleasure is actually saying.

Luke has gotten a bunch of peer-reviewed neuroscience articles and a book he is going through and shared with me. We get together, and we discuss what this means to desirism – a theory of morality based on desires.

Okay . . . here’s what the research seems to be telling us.

The neuroscience seems to be supporting a distinction between ‘wanting’ and ‘liking’. That is to say, the neuroscience seems to be telling us that we can want something without liking it.

In other words, we can be motivated to pursue some state of affairs without there being some type of “hedonic gloss” – that is the term used in the literature – on what it is we are pursuing. We can want things without pleasure.

Well, that’s what desirism has been saying, that a desire is a disposition to act to bring about some state of affairs and is not necessarily associated with pleasure. The neurscientists are saying that we can want without the object of our want having a hedonic gloss that, when present, turns “wanting” into “liking”.

Right now, Luke’s looking at the literature in a bit more detail trying to figure out if we have this interpretation correct and it has the implications we think it has. Then, we will write it into the podcast.

We are also debating whether there is enough interest and relevance to the neuroscience of pleasure to make it a separate episode.

I think there is. I think presenting a theory of morality called “desirism” would benefit by presenting at least a rough sketch of what the neuroscientists are saying about pleasure, and about wanting without liking.

I’m certain when this is done, Luke is going to make me entirely rewrite episode 14 . . . again. I’m also going to be writing a draft of an episode on the neurscience of pleasure and desire – though we may or may not use it.

This has been a very good project for me. I am catching up on a lot of stuff that I have not paid particularly close attention to in the last 20 years.

- Alonzo Fyfe

 

Luke’s Thoughts

My ideas on the use of desirism, my philosophy of language, and my understanding of the neuroscience of desire, have all evolved since Alonzo and I launched Morality in the Real World in September 2010. Luckily, these changes haven’t invalidated any of the basic claims we made in our first 13 episodes.

Most dramatically, my philosophy of language has changed. When originally planning the podcast, I planned on eventually arguing that desirism was the ‘best’ (most useful) set of definitions for moral terms, one that (1) captured common use of the terms and (2) successfully referred to real things to the greatest degree, and that an understanding of morality as an essentially contested concept would allow desirism’s moral reductionism to avoid ‘Twin Earth’-like semantic objections to moral reductionism.

Around the time of the podcast’s launch, however, I lost interest in anything resembling conceptual analysis, and I saw little reason to argue that desirism had the ‘best’ set of definitions. We can always replace the symbol with the substance. Definitions for moral terms aren’t core to desirism; factual claims are. We can propose adoption of desirism’s language by stipulation, but if people want to speak a different language then there’s little I can do to persuade them – unless I can show that their language blatantly fails to refer, which it often does.

Also, my growing interest in the problem of machine ethics for superintelligence (aka Friendly AI) helped me to see more clearly how desirism can be applied. Desirism is a theory that integrates metaethics, normative theory, and applied ethics. Its normative theory, as usually described, depends on the adoption of its set of definitions for moral terms (as is the case for all moral theories). For example, by desirism’s definitions, it’s quite possible (even common) for an agent to not have a reason to do what is ‘morally good.’ But this linguistic proposal is not essential to desirism, and is most useful in the context of human systems of morality.

You can still use desirism to answer such an alien question as “How should we design the motivations of a superintelligent machine?” (This is the problem of SAMA design.) But there are many ways to interpret this question about what we “should” do. You might be asking “For which SAMA design are there the most reasons for action to implement?” Or you might be asking, “Which SAMA design do I have the most reason to implement?” Or you might be asking, “Which SAMA design do existing humans have the most reason to implement?” Desirism has a linguistic proposal about how to interpret ‘morally good,’ but that shouldn’t distract you from the underlying facts we’re trying to shortcut our way to with linguistic proposals.

All three ‘should’ questions above are in principle answerable questions, according to desirism, but desirism can’t tell you which one you ‘really’ should be asking. If you’ve answered the above questions, no question remains about what you ‘really’ should be taking into account. It’s like answering all the questions about Pluto’s mass and orbit and density and makeup and shape and history and then asking “But is it really a planet?” The word ‘planet’ is just a shortcut we made up for talking about a certain cluster of features. If we’ve identified all those features already, then no question remains about whether it’s ‘really’ a planet. It depends which definition of ‘planet’ you mean. And the same goes for the word ‘should.’ It depends which definition of ‘should’ you mean, and if you answer all the questions about possible definitions of ‘should’ directly, then no question remains about what we ‘really’ should do. I know your intuition feels like there should be more, here, but that’s because that’s how the cognitive algorithm that generates your intuition feels from the inside.

Also, my studies of the neuroscience of pleasure and desire have for the most part confirmed my Humean views on how motivation works in humans. Though, the human brain is a ridiculous kluge, so a finished neuroscience almost certainly won’t be as clear as something a philosopher can state in three paragraphs.

My studies of artificial intelligence (that is, of minds in general) lend even stronger support to the Humean view. (See, for example, chapters 2 and 16 of AI: A Modern Approach.)

Right now I’m slogging through a bunch of the literature on the neuroscience of pleasure, desire, reward, learning, goal-directed behavior, and mental representation. We know thousands of very specific things, but we don’t understand very well the picture of those things work together. The way the brain is designed is absurd. Stupid design, indeed!

- Luke Muehlhauser

Previous post:

Next post:

{ 27 comments… read them below or add one }

Martin Freedman March 29, 2011 at 5:58 am

Kep up the good work guys

Having originally started out as a cognitive scientist years ago, this made it very easy for me to not only understand but recognize the strength of desirism as an empirical model of morality. Indeed I hypothesize that any up to date real practioner in cognitive science (as opposed to fan boys of one or another of the philosophers of AI e.g. Searle, Dennet or Chalmers) will grok it more easily than most. So I am both pleased that you both are more closely exploring the neuro-scientific basis of desires, which, whilst my neuro-sciencitific expertise is out of date, I do try and keep up with the literature – I am very dubious of a Singularity – regardless I think you will find neuro-science confirms the key assumptions of desirism rather well.

I agree with you, Luke, on the problems with conceptual analysis within philosophy and am annoyed that I allowed myself to get side-tracked into this too much and particularly philosophy of religion – which I long ago pointed out to you I thought was a boring and not exciting topic – ironically, it was your site that got me into it!

The reason being, whilst focused on ethics, I thought it only fair in granting, for the purposes of argument, that a god exists, in order to show the incoherence of theistic morality. It might be true, I am sure it is regarding Divine Command Ethics but is a waste of time if there is no evidence for such a god and much evidence against and I now understand much better why Alonzo has long avoided going down that route.

So it is great that you have moved on to machine ethics and more than a little bit frustrating than I am too busy to contribute even from the peanut gallery and that I did not focus myself on this two or three years ago. (Of course, that is why I never discussed mind or consciousness – I might never have used those terms as they are not required for moral agency to work – although “imagination” as in the capacity to simulate alternatives and weight the outcomes is – but that has already been long done in AI, see Aleksander’s work on Magnus)..

Finally I think we all agree that desirism can be considered as having a neo-Humean and realist basis, contrary to the popular conception of Hume. Indeed, I have two academic books arguing for a realist reading of Hume, so you are in good company (I have had no time to read them yet though).

Looking forward to future podcasts and I would say, it is better to go more thoroughly and so please do make an episode on the neuro-science of pleasure and desire. I think that this can only help in getting the attention of practioners from that domain interested, if only sceptically (but there is nothing wrong with that) in desirism.

  (Quote)

Jordan Peacock March 29, 2011 at 7:03 am

Thank you for taking your time with this to do it right. I applaud your rigor.

  (Quote)

PDH March 29, 2011 at 7:39 am

It seems like you’ve hit on some really productive avenues of research.

  (Quote)

Luke Muehlhauser March 29, 2011 at 10:10 am

Thanks, Martin!

  (Quote)

cl March 29, 2011 at 11:49 am

I applaud the both of you, too. This seems like a far more fruitful approach than the “accept what I say without any evidence” approach that’s characterized the discussion thus far. Now that desirism seems to have at least potentially moved beyond the, “Fyfe doesn’t like X therefore we have reason to promote an aversion to X” stage, I’m more optimistic.

  (Quote)

mpg March 29, 2011 at 12:27 pm

I applaud the both of you, too. This seems like a far more fruitful approach than the “accept what I say without any evidence” approach that’s characterized the discussion thus far. Now that desirism seems to have at least potentially moved beyond the, “Fyfe doesn’t like X therefore we have reason to promote an aversion to X” stage, I’m more optimistic.

Wow, cl. You are like a dog with a bone. I just knew you would comment on this thread when I saw it.

  (Quote)

Ralph March 29, 2011 at 1:10 pm

Wow, cl. You are like a dog with a bone. I just knew you would comment on this thread when I saw it.

I thought the same thing. cl, give it a rest.

  (Quote)

mpg March 29, 2011 at 1:14 pm

I thought the same thing.cl, give it a rest.

Don’t want to give the wrong impression. I think cl is quite right in his criticisms of modal skepticism and much of Luke’s reasoning. My criticism is the slightly pedantic pattern he has fallen into. Don’t want to shut the guy down at all.

  (Quote)

cl March 29, 2011 at 6:01 pm

mpg,

I think cl is quite right in his criticisms of modal skepticism and much of Luke’s reasoning.

First I want to say “thanks” for not being one-sided. I find that refreshing.

So, let me get this straight: I come over here, see something that actually counts as a new direction in the discussion on desirism, leave an optimistic comment in support of this new direction, followed by a brief explanation of what I think the previous problem was, and that’s being a “dog with a bone?” I don’t get it.

Of course, in the context of the meta-debate that’s been going on for well over a year now, there is a grain of truth in your criticism: yeah, I’ve been persistently criticizing Luke and Alonzo throughout this discussion. You are more than entitled to your opinion that such is “annoying,” or “repetitive,” or “trollish,” or fill in the blank with whatever adjective you like, yet, truth be told, this entire time, I’ve merely been trying to hold these guys accountable to the standards they themselves set for rigorous inquiry about real-world issues. If they would, oh, I don’t know… actually concede their breaches of said standards–or show how they’ve not breached them–then I wouldn’t need to keep repeating myself.

I take serious issue with special pleading, and we’re all guilty of it from time to time. I mean, Alonzo comes out and claims that that “any morally responsible speaker says, ‘The evidence so far suggests to me that ‘P’, but, of course, I could be wrong,’” yet he also says, “There is no god. God does not exist.” Is that–or is that not–a blatant breach of the standard?

And nobody, except for me, says a thing about it. Doesn’t that strike you as odd? We’ve got all these ostensible skeptics and freethinkers, yet, nobody cares. Should I just turn the other cheek and give up any hope of an explanation? I’m not asking a rhetorical question. I’m asking what you think a person with a strong desire to get to the truth ought to do.

  (Quote)

Adito March 29, 2011 at 6:48 pm

Cl, people are under no obligation to endlessly qualify their statements so long as they state somewhere that the strength of their belief is in proportion to the evidence available. It has been made abundantly clear by both Luke and Alonzo that they hold to this idea so you have no place to criticize them about their lack of qualifications.

As for the rest of your post… I’ve seen them respond to you before and I believe they’ve stopped because you have not offered any interesting objections. If your point about how they should qualify their beliefs constantly is any indication of the strength of your objections then I do not blame them for this.

  (Quote)

cl March 29, 2011 at 7:13 pm

Adito,

It has been made abundantly clear by both Luke and Alonzo that they hold to this idea so you have no place to criticize them about their lack of qualifications.

What I’m hearing from you is something like, “It’s okay to do ~X so long as you state that you believe X.” If that’s not what you’re saying, please clarify.

I’ve seen them respond to you before and I believe they’ve stopped because you have not offered any interesting objections.

Yet, many people believe I offer very interesting objections. Who’s opinion should I accept?

  (Quote)

Esteban R. (Formerly Steven R.) March 29, 2011 at 8:57 pm

Luke, here’s the thing you said that caught my attention:

“All three questions ‘should’ questions above are in principle answerable questions, according to desirism, but desirism can’t tell you which one you ‘really’ should be asking.”

But as far as I understand, subjectivist morality holds something akin to “to a cat, eating a mouse is perfectly fine; to a mouse, being killed by a cat is a moral abomination”. To me, that’s what morality should ultimately do, tell us which questions to ask. Otherwise, we can list so many goals (say, winning a race, eating cake, etc.) and all the things you need to do to achieve said goals, but if you can’t pick one over another, then what you end up doing really is up to each individual based on their individual tastes, nothing more.

Other than that, very interesting read as usual.

  (Quote)

E.G March 30, 2011 at 2:57 am

“The way the brain is designed is absurd. Stupid design, indeed!”

And yet, we still dare to think with it! Interesting…

  (Quote)

TaiChi March 30, 2011 at 3:35 am

“Around the time of the podcast’s launch, however, I lost interest in anything resembling conceptual analysis, and I saw little reason to argue that desirism had the ‘best’ set of definitions. We can always replace the symbol with the substance. Definitions for moral terms aren’t core to desirism; factual claims are. We can propose adoption of desirism’s language by stipulation, but if people want to speak a different language then there’s little I can do to persuade them – unless I can show that their language blatantly fails to refer, which it often does.”

If definitions, or at least explications of moral terms aren’t core to Desirism, then it is difficult to see how it can be claimed to be a moral theory. Perhaps, like the analysis of moral terminology, you don’t see much point in arguing for this label either. But what then is Desirism a theory of? What is core to Desirism, which Desirism deserves to be named a theory of?

(A proposal: Desirism is to ethics as eliminative materialism is to the philosophy of mind. Neither theory attempts to analyze the terms with which its field is traditionally concerned with, but each instead proposes a shift to a scientifically-based discussion which covers the approximate area of concern).

  (Quote)

mojo.rhythm March 30, 2011 at 4:13 am

Ha!

You guys fell for Cl’s bait; hook, line and sinker!

I saw that coming a mile away :)

  (Quote)

Martin Freedman March 30, 2011 at 7:53 am

Hi TaiChi

If definitions, or at least explications of moral terms aren’t core to Desirism, then it is difficult to see how it can be claimed to be a moral theory.

The implication here is that any moral theory is a theory of language first and foremost? I don’t think you mean that do you?

I have always argued and, I think, Luke and Alonzo would agree (the latter probably said it first) , that moral language is optional and redundant. Indeed the whole point of a perfectly respectable and, IMV, the most thriving area in ethics in the 21st century is reductive ethical naturalism, which desirism can be classified in.

Perhaps, like the analysis of moral terminology, you don’t see much point in arguing for this label either. But what then is Desirism a theory of? What is core to Desirism, which Desirism deserves to be named a theory of?

As I understand it, an eliminative naturalistic ethical theory would say that these terms are an artefact of folk morality and the entities to which they refer do not exist, this is not just a ontological reduction but an ontological elimination. Still that already is other positions in ethics, such as various anti-realisms, error theory and non-cognitivism.

Now Myself, Luke and Alonzo, following Mackie and others, have long argued using the examples of atoms and planets etc. that one can revise language rather than reject it and invent new terms . Either will in fact do. Further in meta-ethics this is the only game in town, showing what these can and really mean.

So, unless I have deeply misunderstood you, AFAICS your complaint has not merit. However, all said and done, I do like the following:

(A proposal: Desirism is to ethics as eliminative materialism is to the philosophy of mind. Neither theory attempts to analyze the terms with which its field is traditionally concerned with, but each instead proposes a shift to a scientifically-based discussion which covers the approximate area of concern).

I like it in the sense -more inspired by than directly derived from the above – that I do think that one can classify desirism within the conventional meta-ethical framework only with difficulty since it both agrees and disagrees with – clearly different aspects of (all prefixed with “moral”)- non-cognitivism, realism, subjectivism and relativism only excluding absolutism. Still this probably applies to all reductive naturalistic theories. Dunno about Luke and Alonzo, I have made arguments that this conventional meta-ethical classification is broken and needs to be thrown away. We can replace it with what already exists in rational-empirical-historical enquiry and there is no need for a special ethical classification framework.

And note that desirism is also a normative and applied ethics theory, not just a meta-ethical model.

Does this help in any way?

  (Quote)

Martin Freedman March 30, 2011 at 8:11 am

Hiya Estaban R. (and Luke)

Luke, here’s the thing you said that caught my attention:
“All three questions ‘should’ questions above are in principle answerable questions, according to desirism, but desirism can’t tell you which one you ‘really’ should be asking.”
But as far as I understand, subjectivist morality holds something akin to “to a cat, eating a mouse is perfectly fine; to a mouse, being killed by a cat is a moral abomination”.

Desirism is, not, in the sense I think you use it, a subjectivist morality. Rather, desirism can be considered, partly the result of retaining the virtues and excluding the flaws of moral subjectivism (as well as other approaches). There is no morality without desires but that does not imply that morality is subjective – the latter conclusion is a hasty generalisation. This is not original to desirism, as it this applies to pretty much all desire-based ethics theories, which, IMV, dominate the area of reductive naturalistic and realist ethics today. But I will stop falling further into that broken framework i just criticised in my previous post!

Replying directly to Luke I think all three of your “shoulds” are incorrect. Desirism would surely ask what do people generally have reason to implement in SAMA? I think your issue is that SAMA itself would become a moral agent and a moral patient and so also worthy of consideration – not just people generally? This would imply that praise, blame, reward and punishment must also be applicable to SAMA and, if so, SAMA must be designed to ensure it cannot avoid the equivalent of those social forces on it? The interesting question, I think, is, of course, how to design it so and we can learn much from that task about morality.

  (Quote)

TaiChi March 30, 2011 at 8:14 pm

Hi Martin,

“The implication here is that any moral theory is a theory of language first and foremost? I don’t think you mean that do you?”

No. I don’t really see how that could be: a theory of morality may say what moral terms refer to, but a theory of language would say more, since it would tell us what it means to refer in general.
In any case, my point is more like this: if X is a theory of Y, then there must be something central to theory X which marks it as a theory of Y (in virtue of which it is correct to say it is a theory of Y). Usually, that something involves the analysis or explication of the Y vocabulary. Yet Luke’s remarks indicate that analysis or explication of moral terminology is not central to Desirsm and so this does not mark Desirism as a moral theory. As I don’t know what else could mark Desirism as a moral theory (suggestions, please?), it looks to me that it isn’t one.

“I have always argued and, I think, Luke and Alonzo would agree (the latter probably said it first) , that moral language is optional and redundant.”

Well, I’m just not sure what that claim amounts to. If it’s the claim that we don’t have to use moral language, then I think that’s trivial. If it’s the claim that we can cover most of what we want to say about morality by using an alternative but non-equivalent discourse, then that sounds very much like what the eliminative materialists say about folk psychology.

“Still that already is other positions in ethics, such as various anti-realisms, error theory and non-cognitivism.”

Neither error-theory nor non-cognitivism are theories of morality; instead, these are meta-ethical theories. So perhaps Desirism is a meta-ethical theory: well and good, but that doesn’t make it a moral theory.
Remember too that Luke and Alonzo (at least originally) touted Desirism as a form of moral realism. The fact (if it is one) that Desirism is a form of moral eliminativism is significant: it would be a complete turn-around. Now, I don’t know what they want to say about Desirism these days, but it seems to me that Luke cannot eschew the task of showing that Desirsm provides the correct analysis of moral terminology if he intends to develop the theory along its original lines.

“And note that desirism is also a normative and applied ethics theory, not just a meta-ethical model.”

If it’s a normative and applied ethics theory, then it’d have to be a form of moral realism.

  (Quote)

Martin Freedman March 31, 2011 at 1:23 am

Hi TaiChi as usual some good insights, but I always have high expectations of your thoughts.

I can only speak for my own views here, I am sure Luke can jump in if his is that different.

In any case, my point is more like this: if X is a theory of Y, then there must be something central to theory X which marks it as a theory of Y (in virtue of which it is correct to say it is a theory of Y). Usually, that something involves the analysis or explication of the Y vocabulary.

Yes but, to borrow from Korzybski his only insight that is of much use, lets view the Y vocabulary (YV) as the “map” and what it refers to as “territory”. Desirism directly explains the territory and so shows how the old map of YC has mistakes over what the territory is and that a revised YV can refer to it. That is why YV is both optional and redundant as the territory still exists regardless.

Now the complication for this particular Y is that, unusually, YV has dual meanings, I just very briefly dealt with the usual descriptive aspect. Whilst YV can be used as a short-cut to address that territory (as can any other Y’ specific language for its territory) here, in addition, YV is also used non-cognitively and can constitute typically part of the process of social praise, condemnation, reward and punishment that Y theory must address. Desirism X explains this other aspect of Y too.

In this latter sense YV itself is part of the territory but it is not necessary, one can get by without it – without using good, evil, ought and so on. (Desirism provides substitutable co-referrents that preserve intentionality – but only of the descriptive aspect only, these substitutions do not provide the same illocutionary force but that is to be expected).

Yet Luke’s remarks indicate that analysis or explication of moral terminology is not central to Desirsm and so this does not mark Desirism as a moral theory. As I don’t know what else could mark Desirism as a moral theory (suggestions, please?), it looks to me that it isn’t one.

Deisirsm starts with an explication of the moral language to see what it could possibly mean in the real world, the method used could have found nothing but, as it happens, it did not. So it goes beyond the map to the territory and so leaves that map behind to draw a better map, as should any moral theory. (If there were no moral language the starting point would be different but then one would need to explain why there is no moral language – we do not live in that world – still that is something that SAMA design needs to investigate)

If it’s the claim that we can cover most of what we want to say about morality by using an alternative but non-equivalent discourse, then that sounds very much like what the eliminative materialists say about folk psychology.

Yes we (and we are not alone) are eliminating many illusionary and projective constructs that are mistakenly regarded as the territory of morality – intrinsic prescriptivity, prescriptive law, categorical imperatives, essential goodness and evil and so on, desirism shows that what remains is real – factual relational value – and what the YV can be shown to refer to in practice, regardless of what people think they mean when they are making moral claims.

However your are equivocating over the meaning of “eliminate”. We are eliminating fictional entities from the discourse – we surely agree on that – and we are also reducing it to factual entities (relational values).It appears that you are focused on the first clause and ignoring the second clause? AFAICS that is the only way can you map it to Eliminative theories in cogsci.

To be clear, unlike the eliminativists, we are not saying that what results is entirely different (and that is not implied by saying, as I do at least, “optional and redundant”) , since the language can still be used. “You keep using that word. I do not think that it means what you think it means” only applies to its descriptive meaning, the prescriptive usage remains the same.

Neither error-theory nor non-cognitivism are theories of morality; instead, these are meta-ethical theories.

I quite disagree. For example Mackie used his error theory with the consequential revision of the pragmatic meaning of YV as an argument for a normative preference satisfaction. Similarly Hare with his non-cognitive universal prescriptivism did the same ending up with a preference satisfaction utilitarianism. And both get very close to desirism at different points in their development too (Particularly Mackie with his reduction of generic value to a specific type of relational value and Hare with his, inexplicably undeveloped insight, into evaluating ends as means) . In general what meta-ethical theories can do is provide the foundations for a particular normative ethics, (certainly some deny any but the examples you gave are not of that kind!).

So perhaps Desirism is a meta-ethical theory: well and good, but that doesn’t make it a moral theory.

Given my response above you can see why I think your conclusion does not follow. Your conclusion is also performatively contradicted by the use of desirism as a normative and applied theory as seen in hundreds of Alonzo’s essays and short commentaries.

Remember too that Luke and Alonzo (at least originally) touted Desirism as a form of moral realism.

It was and remains that AFAICS. Importantly I sought an empirically realist approach and found the best working hypothesis in desirism, if you don’t wan to call it a “moral realist” theory makes not the slightest difference to me, as I have said that meta-ethical classification is broken and should be discarded. Desirism still remains an empirically realist theory of morality and that is what matters.

The fact (if it is one) that Desirism is a form of moral eliminativism is significant: it would be a complete turn-around.

It is in the main class of 21st century moral realism, reductive naturalism, where the largest amount of theories are desire-based too. I feel I am beginning to repeat myself. Now if you want to go against the modern understanding of these terms , if I have not made this clear by now, that is idiosyncratically, then go ahead but that makes no difference to what the theory is, a moral realist theory. I think you might have a mistaken understanding of moral realism and reductive naturalism as it is today, that a blog comment is too short to explore. I humbly suggest that you read Railton’s essay Moral Realism available via google. If you disagree with that, then I suggest you create your oww blog post on that – connected with desirism or not – and I will respond to you further argument there.

Now, I don’t know what they want to say about Desirism these days, but it seems to me that Luke cannot eschew the task of showing that Desirsm provides the correct analysis of moral terminology if he intends to develop the theory along its original lines.

All one can show how moral language is actually used pragmatically. This will almost invariably differ from what people think it means folk theoretically, I think what Luke is getting at is if they insist on such mistaken meanings (being mistakes due to invoking fictional entities) well let them. He does not want to waste time on such matters. I agree.

Note I really have little to time to carry this on, although it is always fun and stimulating to correspond with you.

  (Quote)

Martin Freedman March 31, 2011 at 4:41 am

Woops where I said “intentionality” I meant “intensional meaning”

  (Quote)

Esteban R. (Formerly Steven R.) March 31, 2011 at 4:48 pm

Hiya Estaban R. (and Luke)

Desirism is, not, in the sense I think you use it, a subjectivist morality. Rather, desirism can be considered, partly the result of retaining the virtues and excluding the flaws of moral subjectivism (as well as other approaches). There is no morality without desires but that does not imply that morality is subjective – the latter conclusion is a hasty generalisation. This is not original to desirism, as it this applies to pretty much all desire-based ethics theories, which, IMV, dominate the area of reductive naturalistic and realist ethics today. But I will stop falling further into that broken framework i just criticised in my previous post!

Thanks for the response. I’m still trying to understand this morality business as I am very new to the debate and haven’t had any formal studies on this at all. So, right now, I just realized that when I said moral subjectivism I meant moral relativism and that the two terms are not used interchangeably in moral philosophy, or at least don’t seem to. That would explain quite a bit.

I spent a day trying to frame it, and the answer hit me in my government class. My teacher said something to the effect of “Unfortunately for Mexico, the U.S. took California. Or fortunately, if you’re American.” To me, desirism comes off as a form of analyzing the “land” that was lost, if you will. We may note that California is fertile and has water resources. From there, we can conclude that having this piece of territory is favorable within the context of what a Nation-State should have. However, once we introduce competing interests, say Nation-States (or, AI and alien life in the case of desirism), while we may be able to infer what is best for Mexico and the U.S., it does not tell us who is in the right. It just tells us that if we’re the U.S., you should get California and that if you’re Mexico, you should get California.

This seems pretty useless to me. I agree that provided that we have a basis for mutual agreement, desirism works, but it seems more like something based on pragmatism than “morality.”

  (Quote)

TaiChi March 31, 2011 at 5:46 pm

Yes but, to borrow from Korzybski his only insight that is of much use, lets view the Y vocabulary (YV) as the “map” and what it refers to as “territory”. Desirism directly explains the territory and so shows how the old map of YC has mistakes over what the territory is and that a revised YV can refer to it. That is why YV is both optional and redundant as the territory still exists regardless.

I think I get the gist, though I wouldn’t put it that way (in the first sentence you take for granted that the YV refers; in the second, that it fails to do refer). This just looks like my proposal that Desirism is analogous to eliminative materialism put in different words. Yet eliminative materialism is not a theory of mind, and neither (if the analogy is appropriate) would Desirism be a theory of morality. (But read on for clarification regarding that proposal.)

Now the complication for this particular Y is that, unusually, YV has dual meanings, I just very briefly dealt with the usual descriptive aspect.

I take it that you didn’t: the descriptive aspect of YV is left untouched by the recommendation that we should replace it with some other kind of vocabulary. You are no more explaining what the YV describes than the eliminative materialist explains what folk psycology describes. At best, you would give an account of why we are led to use the YV, flawed though it is – but this is obviously a meta-ethical account, and so it does not mark Desirism as a theory of morality per se.

[aside]“(Desirism provides substitutable co-referrents that preserve intentionality – but only of the descriptive aspect only, these substitutions do not provide the same illocutionary force but that is to be expected).

Is it to be expected? The Desirist begins with the claim that morality concerns reasons for action, and one would think that the justification for this claim lies in the fact that moral language does has the perlocutionary* force that it does – that only if the YV refers to reasons for action would we be able to explain its being directive.

* I think this is the word you meant, not illocutionary.[/aside]

Desirism starts with an explication of the moral language to see what it could possibly mean in the real world, the method used could have found nothing but, as it happens, it did not.

Well, here is a suggestion which might mark Desirism as a moral theory. Alas, I think even the bluntest form of moral anti-realism begins with moral language too, and is not a theory of morality.

However your are equivocating over the meaning of “eliminate”. We are eliminating fictional entities from the discourse – we surely agree on that – and we are also reducing it to factual entities (relational values).

I’m afraid I don’t understand: how can you both be eliminating fictional enitities from the YV, and showing that the same entities reduce to factual entities? Surely fictional entities which reduce to factual entities are, in virtue of the reduction, factual entities?
But nevermind, this question isn’t really to the point. My point was that Luke could not eschew the demand to provide an analysis/explication of moral terminology if he wished to maintain that Desirism is a moral theory. That he has such an analysis in hand does not show that such an analysis is not required. Indeed, it suggests the opposite: that the analysis is required, and has been developed because Desirism is supposed to be a moral theory.

I quite disagree. For example Mackie used his error theory with the consequential revision of the pragmatic meaning of YV as an argument for a normative preference satisfaction.

You disagree that error-theory is not a theory of morality because a noted error-theorist has advocated a revised theory of morality, or something like it? That’s a non-sequiter.

In general what meta-ethical theories can do is provide the foundations for a particular normative ethics, (certainly some deny any but the examples you gave are not of that kind!).

Yes, if the meta-ethical theory is moral realism. Not so if we are talking error-theory or non-cognitivism. I leave it to the interpreters of Mackie and Hare to say whether they advocate a conjunction of error-theory with some alternate normative system to morality, or whether they are actually moral realist who adopt an error-theoretical stance for methodological reasons. To take the view you suggest, that they are error-theorists and moral realists too, is to saddle each with a contradiction that no charitable interpreter would contemplate.

Your conclusion is also performatively contradicted by the use of desirism as a normative and applied theory as seen in hundreds of Alonzo’s essays and short commentaries.

Fine, so it’s a form of moral realism. Back on the other horn of the dilemma, then: as a moral theory, Desirism is, after all, required to analyse and/or explicate the YV.
Just to be clear, I’m not wed to the proposal that I’ve made of Desirism as a form of eliminativism (and therefore as error-theory). I only made it because I thought that it was consistent with Luke’s assertion that defining the YV was inessential to Desirism. Arguing against the proposal still leaves you with the original problem for which the proposal was to be a solution to: that of maintaining Desirism as a moral theory when it doesn’t, or needn’t, anchor itself to moral concerns by giving an account of moral language.

It is in the main class of 21st century moral realism, reductive naturalism, where the largest amount of theories are desire-based too. I feel I am beginning to repeat myself.

Hopefully, with my clarification on my proposal, you can see that I’m not really arguing against Desirism’s being a form of moral realism. I am suggesting it, because it fits with some of Luke’s thoughts about what Desirism ought to be.

All one can show how moral language is actually used pragmatically. This will almost invariably differ from what people think it means folk theoretically, I think what Luke is getting at is if they insist on such mistaken meanings (being mistakes due to invoking fictional entities) well let them. He does not want to waste time on such matters. I agree.

I’ve no problem with Luke declining to waste time on dogmatists. What I question is whether he can decline to give an account of moral language and continue to maintain that Desirism is a moral theory. Depending on what drives his rejection of the demand to provide that account, I’d suggest he either (i) revises Desirism along eliminativist lines, allows the dogmatists their views of moral language, but tries to refutes the moral claims they make, or (ii), continues to develop a Desirist account of moral language, but takes the claims of the dogmatists with a heavy dose of salt. He need not reply to every critic in order to continue to believe that the Desirist account of moral language is roughly correct, and certainly does not need to do so to continue to believe that it has the best prospects of any other theory on the market.

Note I really have little to time to carry this on, although it is always fun and stimulating to correspond with you.

That’s perfectly alright, I don’t really have the time either. And, likewise, I find it agreeable to disgree with you. :)

  (Quote)

TaiChi March 31, 2011 at 5:48 pm

Just noticed the new slogan: nice one, Luke!

  (Quote)

Martin Freedman April 1, 2011 at 1:56 am

Luke, like the new slogan too! :-)

Estaban. Nice example. Not much time or space to answer properly but heres some thoughts.

The question is as to whether the USA or Mexico were morally right, both answers you give are partial and not universal. A third alternative would be for California to be independent of both and then freely decide which union to join or not.

A desirist analysis would be based on what people generally have reason to promote or discourage. The question is which people and what does generally mean here (I regard it as time and place transcendent and so not dependent on the actual distribution and actual numbers of people involved). I suspect the desirist answer would be that third alternative, it is just that in history it did not happen.

TaiChi

Luke can defend whatever he needs to. Whether I am arguing the same i am unsure. Lets answer you but I will try and keep it short, Hopefully we don’t lose key understanding in doing so.

Nothing here directly pertains to desirism. It does pertain t the class of theories to which desirism belongs, which reductive naturalism domain of moral realism, reductionism for short in what follows. You seem to be denying that reductive naturalism is moral realism. Look it up it is moral realism par excellence.

To clear up a couple of points:

1. Eliminativism versus reductionism

Continuing the map/territory metaphor.

In morality there are many faulty maps, one of the most common reasons for this being them referring to non-existent entities in the territory.

Reductionism says that by removing those fictional entities, one see what the territory really is. Eliminativism says there is no territory. Reductionism, from that territory, builds a better map whereas eliminativism says there are no maps since there is no territory.

I never said that fictional entities can be reduced to factual ones, Fictional entities are reduced to nothing. They are fictions. The reduction remains of mapping moral terms to facts.

2. Meta-ethics and Normative ethics

You disagree that error-theory is not a theory of morality because a noted error-theorist has advocated a revised theory of morality, or something like it? That’s a non-sequiter.

A purported non-realist meta-ethical theory can be used as a foundation for a normative theory. It is you who defined that as impossible. I gave two example to counter that definition assertion. They derived there normative ethics from their meta-ethics, not me. I have two books on Hume and realism that do the same for Hume.

How is this possible, you might say? Clearly they only eliminate some, not all, aspects of a problematic realism and then develop a normative theory with what is left. It is easy to call them morally realist or anti-realist, it just depends on how one defines moral realism. There are no well defined boundaries for all these meta-ethical categories, one’s subjectivism is another objectivism and so on (apart from the extreme positions of course). Now if you insist that only moral realism can lead to a normative ethics, well then their meta-ethics was not really non-realist! Indeed that is the position in current thinking.

3. Conceptual Analysis

What Luke (and I) do not want to do is just a conceptual analysis of (arguably) faulty maps, such conceptual analysis based on ignoring or disregarding the territory and what actually exists. Or indeed the type of conceptual analysis you seem to be performing here on meta-ethical terms!

It appears implicit you are making an argument of the following form, not that you have made this argument, if one wants to insist that any theory that denies say, categorical imperatives exist and that there only hypothetical imperatives then that is not a moral theory. So what, is my response.

It seems you wish to pursue the type of conceptual analysis that Luke complains about- and here you are using conceptual analysis to defend the use of conceptual analysis?! :-) ;-) :-) Or are you trying to demonstrate why Luke considers this a waste of time ? :-) :-)

Really this nothing to do with desirism per se and that has long been dealt with in the last 25 years in academic ethical thinking. Please if you are genuinely concerned about this issue then read the Railton article, you appear to be arguing from an outdated academic ethical perspective.

  (Quote)

antiplastic April 1, 2011 at 12:31 pm

I applaud you finally making a decisive, explicit rejection of the idea that Fyfism in any way aligns with common uses of moral language after all the wiffle-waffling folks like Richard Wein and others tried to pin you down on. This is not a gloat. But I wonder if you’ve seriously thought through all the implications of picking this horn of the dilemma.

If you abandon even the pretense that your collection of shibboleths (or, “theory”) bears any coherent relation to what anyone else on earth refers to when they talk about ethics, then the honest thing to do is to stop using those terms. Stop talking about “morally good desires”. Stop claiming you are doing “machine ethics”. Stop having “guest posts in ethics” by an “atheist ethicist”. Stop calling yourself a “moral realist”, and stop contrasting your “theory” to moral functionalism or noncognitivism because you and they aren’t even talking about the same subject matter. Stop saying you are “treating morality as an engineering problem”. Can you really taboo your words?

  (Quote)

Luke Muehlhauser April 1, 2011 at 12:40 pm

antiplastic,

It’s not so much that desirism is particularly far from what “most people” mean when they use moral terms, it’s that (1) people themselves are confused when they use moral terms, and (2) no theory can capture what most people when they use moral terms, because people use moral terms to mean a wide variety of things. So this is not a special problem for desirism.

Also, there are many people who talk about moral theory very similarly to how desirism does. Perhaps the claim that makes people want to avoid calling desirism a moral theory is that it does not affirm the existence of internal reasons. But neither do many other moral theories in use, for example mid-career Foot.

And yes, we can taboo our words if we try.

  (Quote)

cl April 8, 2011 at 11:15 pm

Esteban R.,

This seems pretty useless to me. I agree that provided that we have a basis for mutual agreement, desirism works, but it seems more like something based on pragmatism than “morality.”

Exactly. This what I’ve been saying all along. Pragmatism != morality.

antiplastic,

If you abandon even the pretense that your collection of shibboleths (or, “theory”) bears any coherent relation to what anyone else on earth refers to when they talk about ethics, then the honest thing to do is to stop using those terms. Stop talking about “morally good desires”. Stop claiming you are doing “machine ethics”. Stop having “guest posts in ethics” by an “atheist ethicist”. Stop calling yourself a “moral realist”, and stop contrasting your “theory” to moral functionalism or noncognitivism because you and they aren’t even talking about the same subject matter. Stop saying you are “treating morality as an engineering problem”. Can you really taboo your words?

Nail on the head. As usual when it comes to desirism, Luke’s response couldn’t pay the bill. That other theories may suffer similar shortcomings does not absolve he and Alonzo from these charges.

  (Quote)

Leave a Comment