Why do people make theories without predictive power?

Question

I studied neuroscience and during my studies I had a course called "The philosophy of consciousness". There we looked at a theory called Qualia, which infuriated me to no ends. Reading up on it online didn't really help. In the exam we only had to parrot what the teacher said. Now Qualia was defined as repeating the sentence: "What it feels like to X". What ever that exactly means. At one point I asked if the theory is a scientific theory and if it holds any predictive power? To which the professor responded, no the concept of Qualia has no predictive power. Now my question is why do people make theories without predictive power? I mean why is Qualia taught while when I say Blorpalorp is: "What it smells like to X" is not taught at universities. More generally why do people create theories without predictive power?

Answer 1

I said this in a previous post

There is a powerful tendency for people in science and computing to think there is nothing very interesting or special about human minds. And unfortunately, a powerful strand in philosophy which says there is something so special about them, scientists aren't on track to figuring them out - the 'qualia' idea and the so called Hard Problem Of Consciousness. I strongly recommend not joining either camp. The story of physics has been from thinking we were a few results away from explaining everything in 1900, and now we don't know what 95% of the universe is made of - our greatest progress has been to begin understanding the scope of our ignorance. I feel strongly we are on a similar trajectory about intelligence (from What is intelligence?)

The way qualia gets used for this, to say there is something deeply or even fundamentally unknowable about minds & subjectivity, is absolutely frustrating, I agree. Personally I think we can know a lot about being a bat, and what we can't know likely isn't worth knowing.

I chase the stubbornness many philosophy people have about qualia to two roots. One is not having a good account of subjectivity and how it arises. And their fallback position, that each creatures unique personal history of experience, gives a special magic flavour to things no one else can experience. I see both as very flawed.

Subjectivity can be accounted for to my mind, with global workspace theory + strange loops. That is, consciousness with subjectivity, is functional, it is where we integrate different into something unified, and self-awareness or self-consciousness is about being able to put an abstraction of ourselves into the workspace's picture of the future, and then feedback loops (eg if I do this x happens, so I should be this kind of person & more x will happen, I am that kind of person), and tangled hierarchies happen (I think of this point as saying, we don't have foundational truths, we start where we are and use lots of approaches meshed together & checking each other, to get consilience).

The unique personal flavour of experience, I see as made impossible by the Private Language argument, which comes out of Wittgenstein's work. The example I like to use comes from the existence of tetrachromatism and colour blindness. Undoubtedly, these people don't see the same 'red' people with normal vision see. Is it 'unkowable' to us? No. There are a spectrum of ways to know, from asking, to adding/subtracting inputs from the added or missing cone to our brains. More importantly, does it matter that their sensory experience is different? This is where I'd point to the Private Language argument to say no, it isn't. Private, incommunicable qualities of experience may exist, but if they cannot be abstracted, communicated, then they cannot be meshed into our private experiences either - into the realm of refinement and acuity we get from discussing and comparing notes, which is what makes us human, rather than wolf-children (when raised without human contact beyond a certain age, humans never gain language). This is rooted in intersubjectivity (which we can get insight into the human flavour of from Dunbar's Number, if you're interested).

Is a bat different? Or an octopus? I would say no. They have subjectivity for the same reason we do. They can adjust and adapt in the same way eg to sensory damage, or brain rewiring following neural damage. And I would say there is intersubjectivity in the same way, but from evolution, rather than honed & refined by language & neocortex. An octopus has deep insights into what it's different prey and predators see, and think. That seems to be why these molluscs have become so intelligent, on this very different evolutionary branch to us. Intelligibility, and communicability, are fundamentally related, because our awareness of ourself is driven by awareness of 'if I was them', by intersubjectivity.

So that's out of the way. Why do philosophers play this game about qualia, and make unfalsifiable ideas without predictive power? It's like the free will debate (which I give my rundown of here of why it's a pseudo-problem Testing Free Will): badly understood definitions creating problems that aren't problems. Wittgenstein described philosophy as therapy, as exactly for unravelling such knots:

“Philosophy is a battle against the bewitchment of our intelligence by means of language.”

I feel qualia is a useful challenge, to those who "think there is nothing very interesting or special about human minds". What and why we experience things is complex, sophisticated, and important to think about. Like many (all?) machine intelligence researchers, I'd recommend bypassing the qualia debate. But, do so with a healthy respect for the simpler deeper mystery, of how noises from our faces make pictures appear in the minds of others. It is not an extension of pointing, but happens within a whole framework of building shared experience - unexamined assumptions and definitions included. And those are where we need philosophy, in order to have a future of thinking better, together.

Apologies for a long post. But you asked a tricky question. Have tried to keep jargon to a minimum.

Answer 2

Why not? There are many interesting propositions that don't make predictions taken on their own, and philosophers are interested in knowing the truth of these propositions (does God exist? Do universals exist? What is truth/knowledge/goodness/beauty? etc). One interesting question is, 'is everything physical?' and with qualia there's something literally staring us in the face that says 'no' (so some would argue). Feel free to join the debate if you're interested, but if you do, no matter what side your on you'll find yourself very quickly making statements that don't lead to predictions. Ex: Maybe you think it's not possible to confidently know the truth of propositions that don't make predictions. In response: 1. Does that make predictions? And 2. I bet I can think of all sorts of things you rationally believe that don't make predictions (or at least, you're rational in believing them whether or not you've bothered to even think about their predictions).

Answer 3

People are interested in how the world really is. The problem is, what really is might not be observable or measurable, but you still want to make theories or models about how it could be, to be able to express ideas.

So, of course Qualia and similar models has no predictive power, because by its very nature it deals with unmeasurable things and as such can't be called scientific theory. But science doesn't deal with every important aspect of reality.

Not being a scientific theory doesn't mean a thing has no worth. Plenty of things are not scientific in nature, simply because they are not quantitatively measurable. Ethics, for example.

We know we experience something. That is Qualia. We know electrical signals are running in our brain. That is quantitatively measurable. The two are most likely connected but we don't know how. Is how I perceive blue the same as how someone else perceives blue? Does salt qualitatively taste the same to every person? I think there will always remain some doubt about this, but it's difficult to talk about this without a concept of Qualia.

Answer 4

Allow me to use a visceral example. Let's say that I have a bad day at the office, where everything just went sideways in unpleasant detail. I decide that the best way for me to get rid of all that pent up negative energy is to punch you, really, really hard. Let's say further (for the sake of the argument) that you don't want that to happen. So you try to convince me not to punch you by telling me that punching you hurts.

Of course, I'm a philosopher — which probably explains why I had such a bad day at the office — so I'm likely to object to that, pointing out that that in fact no: when I punch you it doesn't hurt much at all. Maybe my knuckles hurt a little, but that little bit of pain is well worth the cathartic release the act. You'll quickly see my error and clarify that when I punch you, it hurts you. But then we're in a bit of a strange place: how do I know it hurts you when I punch you? That is a seriously difficult question to answer.

You can apply all the neuroscience you like here: trace the electrochemical reactions that propagate from the point of impact up the nervous system; look at the patterns of neural firing in various cortexes, and compare that to other brain states, etc. But all of these objective measures are data points, or lines on a graph, or squiggles on a monitor. None of them say 'pain' except by association with the subjective experience of pain. In fact — and assuming I want to be stubbornly skeptical — the only way for you to make your case is for you to hit me really, really hard, then show me all the lines and squiggles the machines produce from my brain, so that I can associate the objective measure with my subjective experience of pain. Then I might be willing to say: "Ok, your brain does the same squiggles, therefore you may have the same subjective experience that I do." But I still don't experience your pain; I'm merely willing to accept the idea that you might feel something similar to my pain.

This is where the 'qualia' concept enters the scene. Qualia refers to that internal subjective experiencing that we cannot share with others except by reflection and relation (aka 'sympathetic understanding'). It's an important area of study because much of our social world is built on the coordination of qualia between individuals. Societies work because we share certain basic understandings about what others experience and want to experience. Without that, we would be a society of sociopaths — people incapable of understanding that the inner lives of others are 'real' — and a society of sociopaths isn't really a society at all. Most of the academic writing you'll find on qualia is dry and trite (musings about color or trivialities), but that subjective/objective split is central to human life, and nowhere near resolved in philosophy or neuroscience.

Answer 5

You are infuriated because the concept of qualia is not really clearly explained to you. It is not "what it feels like to X".

Qualia is in fact what is commonly mentioned as "consciousness" or "soul". Qualia is your subjective experience. Yet it has nothing to do with consciousness, in neuroscientific terms.

Let me elaborate on my notion that it is "your subjective experience" a description which I borrow from Wikipedia "Qualia" article:

I am now breathing, feeling the pain of my legs getting numb and my eyes getting dry from sitting in front of my computer screen too long, and that I'm thinking hard to explain to you this concept. All this experience I have at the moment is my qualia. Whereas what you are experiencing from your subjective perspective is yours. It is not a theory, it is probably been described you in a faulty manner.

There is a clear distinction, an invisible border between your qualia and mine, which will never be accessible unless we become telepaths. You may have heard the common proposition that "my red is not your red". It is true, in that we both can access the particular object that evokes the sensation of what is red. However neither I can access your mind's "vision" or "experience" of what is red. This is where it makes sense.

When we both say that we sense or perceive "X" in front of both of us, we actually refer to the object that evokes the same or similar feeling of what is "X", but we can never explain each other what the experience actually "feels". Another definition rises here. This is called explanatory gap.

If I try to explain in a somewhat mathematical language:

Explanatory gap = Subject 1's qualia - Subject 2's qualia.

Qualia does not require to have a predictive power as it is a concept that is not purely scientific, but is trans-disciplinary and is not a theory or hypothesis.

Answer 6

A more interesting question for me is where does "predictive power" actually come from for a theory or merely a concept?

Most problems of a particular field may have many or endless possible explanations and theories, but only those which haven't been invalidated and useful to connect to other theories or can be applied in real world immediately to solve practical issues will be filtered out. Having predictive power is very subjective and should not be a criterion to seek good theories. Once you have a good theory (metaphorical model) which can honestly describe thus effectively explain domain processes, then "predictive power" will almost certainly come as a byproduct when you successfully applies it to the right domain problem. Of course, in reality it may be very complex involving other domains and hard to achieve above idealized result.

Answer 7

I think it's all in the UV title: Philosophy of consciousness.

i.e. (i guess) introducing you to the concepts about consciousness that touch your field of research and you therefore can't afford to ignore because they trace the border, the line beyond which we don't know yet, and a mine for valuable subjects of study. Maybe it was given too much space in your curriculum, I can't know, but it's definitively something valuable to know about.

I guess you were introduced to the hard problem of consciousness, which has been studied by both philosophers and neuroscientists. If you can find a way to approach it scientifically and make a break through, this is Nobel prize material. Something worth knowing about.