Tuesday, August 23, 2016

Who knows best? Personal Happiness and the Search for a Good Life

I would like to be happier. I would like to live a good life. But I often get it wrong. Once upon a time I thought that getting a PhD would make me happy. It didn’t. It made me painfully aware of my own ignorance and more anxious about the future. Another time I thought that going on holidays to Spain for a week would make me happy: what could be better than a week relaxing in the sunshine, without a care in the world? Surely it would be just the balm that my overactive mind needed? But it didn’t make me happy either. It was too hot and I quickly got bored. By the end of the week I was itching to get home.

It turns out that my unfortunate tangles with happiness are not that uncommon. Daniel Gilbert wrote an entire book about these problems called Stumbling Upon Happiness. In it, he argued that various psychological biases mean that humans are not good predictors of what will make them happy. They often stumble: intentionally doing things that make them miserable while accidentally falling into things that make them happy. He suggested that we shouldn’t rely on our own judgments about what will make us happy. Instead, we should rely on others: we should learn from their mistakes and successes.

Some scientists think they can come up with better tools for judging what will make us happy. Allen McConnell and his colleagues, for example, have developed a modified version of the implicit association test (IAT) that can use our implicit preferences to predict what will make us happy. Similarly, Robb Rutledge and his colleagues have developed a computational model of the brain that can predict subjective well-being during certain tasks. These tests and models are in their relative infancy, but they suggest ways in which careful scientific scrutiny of our minds could assist in the search for happiness.* All of which provokes the following question:

Who knows best? - When it comes to figuring out what makes us happy, who knows best: (i) ourselves or (ii) scientists who have carefully studied the neural and cognitive markers of happiness?

That’s the question asked in Stephanie Hare and Nicole Vincent’s article “Happiness, Cerebroscopes and Incorrigibility: Prospects for Neuroeudaimonia”. They pose an interesting thought experiment. Imagine if some future scientific discoveries allow us to construct a cerebroscope, i.e. a device for looking at the activities and networks within our own brains and identifying the patterns that are correlated with happiness. Should we rely on the cerebroscope in lieu of our own subjective judgment?

They make three arguments in response. First, they suggest that you have to distinguish between two different versions of the ‘who knows best?’ question. Second, they suggest that on the first interpretation of the question, we will always be more reliable judges of our own happiness than any scientist or prospective cerebroscope might be. And third, on the second interpretation of the question, they suggest that scientists and prospective cerebroscopes might be able to offer some useful assistance, but we probably shouldn’t overestimate their contribution.

I want to look at all three arguments in what follows. While I agree with much of what Hare and Vincent have to say, I think their second argument is less important than they seem to believe and that their third argument underestimates the prospects for future technological happiness-assistance.

1. Two different versions of the ‘Who Knows Best?’ Question
Hare and Vincent’s first argument is not really an argument so much as it is an observation. There is little to disagree with in it. But it is important as it sets the stage for their two other arguments.

We must start by noting that ‘happiness’ is complex concept. It’s one of those terms that bandied about a lot in philosophy and psychology and is open to different interpretations. Hare and Vincent focus largely on mental understandings of happiness (i.e. understandings that suppose happiness to be a mental phenomenon). They do this because they are interested in the potential for the mind sciences to contribute to our search for happiness. But they acknowledge that happiness may not be a strictly mental phenomenon.

Furthermore, even if it is a mental phenomenon, there is room for disagreement about the nature and character of that mental phenomenon. Happiness might be raw conscious pleasure or euphoria, or it might be a less extreme feeling more akin to satisfaction. Alternatively, happiness could be a diffuse mood or temperament that affects how you perceive and understand the world. At the very least, we can say that it is a positive mental feeling/state/mood and that it can be ephemeral or long-lived.

Humans are interested in happiness because of its positive attributes and because they believe happiness is an important part of a well lived life. Most people want to live happy lives. But then they must confront the two different aspects of happiness that could be relevant to their quest for happiness:

In the moment happiness: This is the occurrent mental state/feeling/mood of being happy.
Dispositional happiness: These are properties or attributes of the individual that make them likely to experience ‘in the moment’ happiness in particular contexts.

Suppose I am an opera-lover. I am endlessly fascinated by the opera; I like to attend operatic-performances as often as I can. This suggests that I am disposed to experience ‘in the moment’ happiness when I attend an operatic performance. But, of course, just because I am disposed to be happy while attending the opera does not mean I will actually experience happiness while there. I could attend a particularly bad performance which upsets my sophisticated tastes.

These two different aspects of happiness have knock-on effects on the ‘who knows best’ question. Indeed, they suggest that there are two parallel versions of that question:

Who knows best?1 - Who knows best about my occurrent, in the moment, feelings of happiness?

Who knows best?2 - Who knows best about my dispositions toward happiness and the contexts and experiences that are likely to make me experience in the moment happiness?

Hare and Vincent suggest two different answers to both of these questions.

2. Who knows best about ‘in the moment’ happiness?
The first version of the question has a simple answer: we do and we always will know best the current state of our happiness. Hare and Vincent defend this answer by reference to Richard Rorty’s idea of incorrigibility. Rorty was interested in finding one of the distinctive marks of the mental. Many have been proposed over the years: intentionality, consciousness, privacy and so on. He thought incorrigibility was a mark of the mental.

By ‘incorrigibility’ he meant that certain self-reports about mental activity are incapable of being corrected. For example, when you say that are experiencing the colour red in your visual field right now, your self-report is incapable of being corrected. I might say to you ‘but there is nothing red in your visual field right now’, and objectively speaking I might be right, but that doesn’t mean that you are wrong about your subjective experience. Or, to take another example, suppose I claim that right now I am thinking about the movie I watched last night and how bad it was. Would you be able to jump in and say ‘No, you are not thinking about that’? Of course not. I am the only real authority on the current contents of my conscious thought. Claims about occurrent subjective experience are, in this sense, incorrigible.

Something similar must be true for in the moment happiness. If someone says that they are currently experiencing a state of happiness, who are we to correct them? If my friend turns to me at his mother’s funeral and says that he is feeling happy, I might proffer the view that it is inappropriate for him to feel happy at this time, or predict that the happiness is a strange psychological reaction to trauma and that it will soon pass, but I cannot doubt that he is genuinely experiencing happiness. Judgments of in the moment happiness are incorrigible. The subject of the experiences always knows best. No amount of scientific discovery could disrupt this.

This seems right but it is not beyond criticism. Rorty himself suggested that the incorrigibility of ‘in the moment’ judgments was contingent on current technology and that if a cerebroscope was invented that allowed us to see the current state of our brain activity we might have reason to doubt our own judgment. But Hare and Vincent argue that this is wrong. No matter how sophisticated and precise the cerebroscope becomes, your judgment of your own in the moment happiness would always be incorrigible.

They offer three arguments in support of this. First, any hypothetical cerebroscope would have to be built upon a foundation of self-reported judgments of happiness. A scientist would get a subject to report on their current feelings and then correlate these self-reports with brain states. This would enable them to build a model of the subject’s brain that would offer meaningful predictions about whether the subject is currently experiencing happiness. It would not enable them to question the epistemic authority of the subject’s self-report. Their entire scientific project presumes that the self-report of happiness is correct. The model could not be built without that presumption.

This feeds into a second argument. Our knowledge of the human brain is always going to be incomplete. And what we currently know suggests that the brain is remarkably adaptive and flexible. Brain regions that we think are correlated with one particular mental function can be co-opted and used for another function (particularly in cases of disease or damage). So if we did end up in a situation where our cerebroscopic model told us that the subject was unhappy, but the subject insisted that they were happy, this would really be an opportunity to adjust the model to take account of new data, not to question the judgment of the subject.

This brings us to a final argument, which is slightly more philosophical in nature. It is a variant on Frank Jackson’s classic ‘Knowledge Argument’. That argument is based around the famous ‘Mary in the Black and White Room’ thought experiment. I’ve written about it at greater length before. Jackson asked us to imagine Mary - a scientist of human visual perception —who spent her entire life in a black and white room, dressed in black and white clothes, with black gloves and no mirrors. She knew everything a scientist could possibly know about the visual experience of the colour red. One day, she was released from her black and white lab and went out into the real world. She saw a red apple. Here’s the question: would she learn something knew from her experience or would she already know it all thanks to her scientific discoveries? Jackson insisted that she would learn something knew because no amount of third party data about human visual experience could allow her to know what it would be like to experience the colour red. Scientific inquiry is simply incapable of telling us anything about qualia (i.e. the quality of experience). Assuming Jackson is right, no cerebroscope could tell us anything about our in the moment experience of happiness.

Hare and Vincent consider a more nuanced objection to this way of thinking about in the moment happiness. According to some theorists, occurrent happiness is a broader, more diffuse and more complicated thing. We often suppress or ignore aspects of our current affective experiences or moods that might call into question our claims to happiness. For example, somebody who suffers from anxiety might, on occasion, not realise how their anxiety affects their feelings over time. Hare and Vincent think this is plausible insofar as it goes but that it doesn’t affect the incorrigibility of occurrent in the moment judgments. Those judgments are still correct even if they fail to factor in things that might affect happiness over a longer timeframe. Think back to my interaction with my friend at his mother’s funeral. I might argue that he is ignoring certain emotions that might affect his happiness in the near future, but I still cannot question his occurrent judgment as to his own happiness.

For my part, I think Hare and Vincent are right to say that one’s own judgments of one’s own occurrent happiness are incorrigible, but this is quite a narrow claim and isn’t as practically significant as you might think. There are two reasons for this. First, I don’t think anyone doubts their own judgments of their own occurrent experiences. So I don’t think anyone will be looking for help from scientists or others to figure out whether they are occurrently happy. What they will really be interested in is figuring out what kinds of activities or states of being are likely to generate and sustain that occurrent sensation. That’s where they will need help and where scientists might be able to provide it (see below).

Second, the only cases in which we might call into question occurrent judgments are when we are interested in what other people are feeling and we doubt the verisimilitude of their reported judgments. Thus, I will never question my own judgment about my occurrent happiness, but I might question my friend’s. If he says he is happy at his mother’s funeral I might be inclined to think he is lying or putting brave face on it -- that his self-report is not a true reflection of his internal experience. People often deceive and mislead others. In those cases, philosophical argumentation in favour of the incorrigibility of one’s own judgments is unlikely to make a practical difference. Third parties will look for ways to figure out what a person is truly thinking. That’s why there is so much interest in things like brain-based lie detection devices (and other deception detection techniques).

3. Who knows best about dispositional happiness?
The second version of the who knows best question focuses on dispositional happiness. That is to say: it focuses on figuring out what kinds of states and activities are likely to induce occurrent feelings of happiness given our dispositions and traits. It is a predictive question. When answering this question, we want to know where to turn to when making decisions about our lives. For example, I want to know whether writing more blog posts or articles is likely to make me happy or whether I should do something else with my precious time. Should I trust my own judgment on this matter or should I turn to others for help?

Hare and Vincent agree that scientists could have important advice to offer when it comes to this matter. To illustrate, they sketch a story of young woman who is unlucky in love. She always chooses the ‘wrong type of guy’. They suggest that something akin to McConnell’s modified implicit association test, could help her out. She could find out what her implicit preferences are when it comes to romantic partners and use this data to chart a better course through the rocky waters of romance.

But they then suggest that there are three limitations to the advice that scientists can offer.
First, what actually makes us happy is likely to be a multi-factorial phenomenon. Many factors will combine to make you happy and some of those factors will be ignored or downplayed by both yourself and your putative scientific advisor. It’s worth quoting from them in full on this point:

Everyone has blind spots when it comes to predicting the future; this holds true for scientists as well. You can only base your predictions and choices on your incomplete and maybe even incorrect, self-knowledge and past experiences. But, neuroscientists cannot do much better, and arguably the decontextualized brains that they tend to study might even mean that their chances of making the right prediction are lower. After all, they only base their predictions and advice on what they know about your brain. 
(Hare and Vincent 2016, 81)

I am sympathetic to this point of view. Scientific insight has its blindspots and limitations. We should always keep that in mind. But I wonder whether this criticism doesn’t underestimate the potential for technological assistance when it comes to predicting our own happiness. I appreciate this may be an instance of someone with a hammer seeing a nail everywhere he looks, but my current research on big data, surveillance and algorithmic governances makes me wonder whether those technologies could be leveraged to create highly-individualised, constantly updated, predictive models of an individual’s moods. These models could be based on thousands (millions) of datapoints, helping to address the problem of blindspots and multifactorial causes. I then wonder whether we could each have personalised ‘happiness oracles’ who provide ongoing advice on what we should do to make us happy. These oracles would not be individual scientists or people; they would be sophisticated AI assistants, perpetually mining our personal mood data for important correlations and using this to issue recommendations. Interestingly, Hare and Vincent hint at this possibility earlier in their article when they discuss the phenomenon of mood-tracking apps, but they seem to retreat to a narrower conception of scientific assistance later on when highlighting the limitations of third party advice.

They are on firmer ground with their two other limitations. The second is that we should not confuse predictions of future happiness with normative guidance. This is important. For instance, a modified IAT might reveal an implicit preference for sexual relationships with children. This does not, of course, mean that you should have a sexual relationship with a child in order to be happy. What you should or should not do in a particular context depends on more than just a prediction of your likely happiness. This feeds into their third limitation which is that happiness is only one component of human flourishing.

Those points seem unobjectionable to me and should warrant some caution when it comes to weighing the predictive powers of technology in our quest for the good life.

* (note: I stole the examples of McConnell and Rutledge’s tests/models from Hare and Vincent’s article).

No comments:

Post a Comment