Why Sam Harris shouldn’t change his mind

(if all arguments are as bad as Jonathan Haidt's)

There is an ongoing war in Philosophy, in case you haven’t noticed: Realism vs Relativism. Is there an objective reality? Or is the universe –or multiverse– inseparable from the subjectivity of the human mind? An especially fierce part of this war is fought in the realm of Moral Philosophy: Moral Realism or Moral Relativism? Is morality objective or subjective?

Most people who agree with the former, only do so because they believe the source of their morality is some supernatural entity. Most secularists, and even otherwise “realists”, on the other hand, claim that morality has no basis in nature, and therefore is an entirely human construct, devoid of any “objectivity”.

Much recent work in science and philosophy has shaken that widely-held belief. Even though making the case for moral realism per se isn't the purpose of this essay, I'll mention some examples, and you can read more about them yourselves, if you'd like. Dutch primatologist and ethologist Frans de Waal has showed with a life-long career in research that proto-morality, basic empathy and reciprocity, exist in other mammals, and especially so in primates (Frans de waal: Moral behavior in animals). This suggests that morality is simply an emergent phenomenon, naturally selected for and brought about by evolution. (Ironically, however, Frans seems to agree with Haidt’s most recent article).

Harvard University psychologist Steven Pinker in his book “The Better Angels of our Nature” argues that violence has been decreasing throughout human history and considers it unlikely that the explanation lies in changes in human nature. As many other factors of human experience remain unchanged, it’s more likely that human nature is simply characterized both by inclinations toward violence and those that counteract them; and the latter seem to be slowly but surely prevailing (Steven Pinker: The surprising decline in violence). Aren't societies slowly but surely converging towards the same core moral values of peace, equality and fairness?

I'm painfully aware that it was simple chance
I was born in a place and time where my basic
human rights were more or less protected

It could be said, as this is being composed by a strictly middle class, young, Greek, female citizen, literate enough to know how to write, educated enough so she can understand the topic under discussion, “wealthy” enough to own a netbook and an internet access, and free enough to express her opinion, that we’ve come a long way in terms of human rights, which are, of course, closely related to morality. Just a few decades ago (or a few hundred kilometers away) she would be unable or forbidden to participate. Are those "human rights" just a matter of personal taste? 

But maybe the boldest of moves on this war-field has been made by philosopher and neuroscientist Sam Harris. In his book, “The Moral Landscape”, he makes the most comprehensive and concise case of moral realism as of yet: One need only accept that morality has to do with the well-being of conscious creatures. Then, if we imagine a situation where all conscious beings are suffering as much as they can for as long as they can, this is, by default, the worst possible world to live in. Anything else is, by default, better. This immediately shows that there can be objectively better and worse situations to be in, and therefore there can be objectively better or worse answers to moral questions. There are lots of details to be discussed, but the essence is that morality is, at least in principle if not also in practice, objective (Sam Harris: Science can answer moral questions).

My point is not to support Harris’ claim. I do happen to agree with it (and disagree with several of his other positions, by the way), but that hasn’t prevented me from really trying to find a good counter-argument. Exactly because I prefer the idea that morality is objective, I want the disproof presented to me sooner rather than later. Not only have I failed, but any kind of criticism I’ve heard or read keeps repeating the same four or five straw-men, red-herring etc arguments, to my -and apparently Harris’, as well- disappointment. A few months ago, that led Harris to announce “The Moral Landscape Challenge”, challenging anyone to prove that his core argument about moral realism is flawed, for the not-so-modest prize of $10.000.

In an essay, Why Sam Harris is Unlikely to Change his Mind, social psychologist Jonathan Haidt, writer of “The Righteous Mind” and a moral relativist, counters the bet: “The Righteous Mind challenge: If anyone can convince Harris to renounce his views, I’ll pay Harris the $10,000 that it would cost him to do so.”

Just a friendly piece of advice: If anyone is serious about “changing Harris’ mind”, they’d better do a much better job than Haidt does.

In the above essay, Haidt argues that “[...] science has a long track record of undermining claims about God’s role in the material world”, but it has also “undermined claims about the role and reliability of reason in our daily lives”. He goes on to list some of the human cognitive biases science has exposed, one of which is the emotional attachment to ideas; not wanting to change our minds. So his claim is that science, by exposing our biases, has undermined the role and reliability of reason itself. Isn't that just word-play? He's just embedding in the definition of the term “reason” the quality of its application. Reason can be applied correctly or falsely. That is not a short-coming of reason itself. You can use any tool in a wrong way. You can try to wash clothes in a dish washer, you can try to lock a door with a nail polisher. It's not the tool that's failing when you are not using it as you're supposed to. Would we have ever gotten to expose some of our biases if reason was not capable of overcoming them in the first place?

Later on, Haidt confuses us even more by saying: “Reason is indeed crucial for good public policy and a good society. But isn’t the most reasonable approach one that takes seriously the known flaws of human reasoning and tries to work around them?” What does that even mean? Again, it's just word-play. How do we “take seriously the known flaws of human reasoning” if not by using more, better reasoning? Is there any other way of exposing biases than through reasoning? Horoscopes? Gods? Hunches?

Let's ignore that little bit of weird reasoning about reasoning. He goes on: “If reasoning is so easily swayed by passions, then what kind of reasoning should we expect from people who hate religion and love reason? Open-minded, scientific thinking that tries to weigh the evidence on all sides? Or standard lawyerly reasoning that strives to reach a pre-ordained conclusion? When I was doing the research for The Righteous Mind, I read the New Atheist books carefully, and I noticed that several of them sounded angry. I also noticed that they used rhetorical structures suggesting certainty far more often than I was used to in scientific writing – words such as “always” and “never,” as well as phrases such as “there is no doubt that…” and “clearly we must…””

Well, for one thing, these books are not strictly speaking scientific, as they are directed to the general public, but let’s ignore that, too. Notice the two claims Haidt is making in one breath:
a) Using more “rhetorical structures suggesting certainty” means you, personally, are more certain, and
b) Being more certain than someone else is evidence against “open-minded, scientific thinking that tries to weigh the evidence on all sides”, and for “standard lawyerly reasoning that strives to reach a pre-ordained conclusion”.
He doesn't say how he knows this, he simply accepts it as a given (to reach his pre-ordained conclusions?). And, then he claims to have proven it, by following this highly scientific “methodology”.

He basically ran the texts of a few books through a text analysis software program, the Linguistic Inquiry and Word Count, outputting the "Certain category" which contained the following words and word stems:

“absolute, absolutely, accura*, all, altogether, always, apparent, assur*, blatant*, certain*, clear, clearly, commit, commitment*, commits, committ*, complete, completed, completely, completes, confidence, confident, confidently, correct*, defined, definite, definitely, definitive*, directly, distinct*, entire*, essential, ever, every, everybod*, everything*, evident*, exact*, explicit*, extremely, fact, facts, factual*, forever, frankly, fundamental, fundamentalis*, fundamentally, fundamentals, guarant*, implicit*, indeed, inevitab*, infallib*, invariab*, irrefu*, must, mustnt, must’nt, mustn’t, mustve, must’ve, necessar*, never, obvious*, perfect*, positiv*, precis*, proof, prove*, pure*, sure*, total, totally, true, truest, truly, truth*, unambigu*, undeniab*, undoubt*, unquestion*, wholly”

And the results showed that “the New Atheists win the "certainty" competition”. And that means that “Sam Harris is unlikely to change his mind”. Lots and lots of claims, little, if any, evidence.

Now, as a good skeptic, I did my research, read the information on the LIWC site, and even took the liberty of contacting them for further information, just in case I was missing something, as neither English is my native language nor text analysis my field. Roger Booth was kind enough to respond to me:

“LIWC only claims to report the proportion of words in various categories. Beyond that it’s up to the person using the analysis to interpret and, if necessary, draw conclusions or make claims based on the results. In that, interpretation is always comparative - for example, do people who use a higher proportion of certainty words have consistently differently characteristics that people who use a lower proportion.”

I followed up with a few more questions:

Me: “So, to your example, is it an assumption to say that someone who uses more "certainty" words is exhibiting more certainty? From what I've understood, one would have to prove that this is the case and not simply accept it as a given, simply because there is a category in LIWC that's called "certainty"?” 

Roger: “Yes that's correct. Increases in the number of certainty words don't necessarily indicate more certainty just a higher proportion of that category's word use.”

Me: “The syntax and the context aren't taken into consideration at all, correct? So, for example, the sentences "I'm sure" and "I'm not sure" both have one "certainty" word?”

Roger: “No, context isn't taken into account except in the use of two English words "like" and "kind" where they can be scored as affect words or fillers depending on context. However with all other words there is not contextual analysis. So "I'm certain" and "I'm not certain" would both score in the "certainty" category but the second example would also score in the "negations" category as well and so the researcher might be led to some qualification of the category scores based on that.”

Apparently, simply because there is a “certain” category that does not in any way imply a connection between the number of words used and expressing more certainty; that is an assumption. It’s up to you to prove that this connection exists; and as a commentator on Frans de Waal's post about Haidt's essay said: “No statistics, no replication, minimal percent differences. Garbage comparison.” To put it more politely, this might be something, might be nothing, but it certainly is not “proof” of anything; except perhaps of a confirmation bias at work.

First of all, a lot of those words have many uses, many of which have nothing to do with certainty. Off the top of my head, consider the following: “Absolute zero”, “substances essential for life”, “positive magnetic pole”, “commitment to law”, “statistical confidence”, “total eclipse”, “logical truth”, “perfect hair”, and the list could go on and on. But let’s even ignore that.

Second of all, the context was completely ignored. Let’s see some of the problems with an example: “Fundamentalists tend to ignore the facts and confidently claim certainty. How could I, on the other hand, be sure about anything?”

a) There are four “certainty words” in the first sentence, but the sentence itself shows nothing about the certainty of the writer; it’s actually describing the certainty of others, and it doesn’t even do that with certainty (see “tend to”). You could even quote somebody else and his “certainty words” are put in your mouth.

b) The second sentence also has a “certainty word”, but it’s actually used to the opposite effect: “How can I be sure of anything?”. Apparently, there is a mechanism at place that can detect negation. But language is so flexible, that this is hardly enough.

But, just so I’m not accused of creating an example, made to prove my own point, I’ll quote a commentator's response to the original article, with excerpts from Sam Harris’ book “The Moral Landscape” that contain the “offensive” “certainty words”:

“Always:

“Rational, open-ended, honest inquiry has always been the true source of insight into [facts about the well-being of conscious creatures]. Faith, if it is ever right about anything, is right by accident.” 

“Many people seem to think that a universal conception of morality requires that we find moral principles that admit of no exceptions. If, for instance, it is truly wrong to lie, it must always be wrong to lie—and if one can find a single exception, any notion of moral truth must be abandoned.” 

Never: 

“The world’s profusion of foods never tempts us to say that there are no facts[?] to be known about human nutrition or that all culinary styles must be equally healthy in principle.” “And science and religion—being antithetical ways of thinking about the same reality—will never come to terms.” 

That’s last one is pretty dogmatic, sure. But I agree with that statement. Faith that runs contrary to evidence will never reconcile with knowledge based on evidence. Gays either burn in hell or they do not. It can’t be both with the same meaning of the words. 

Certainly: 

“I am certainly[?] not claiming that moral truths exist independent of the experience of conscious beings—like the Platonic Form of the Good—or that certain actions are intrinsically wrong.” 

“Which is to say that there may be some forms of love and happiness that are best served by each of us being specially connected to a subset of humanity. This certainly appears to be descriptively true of us at present.” 

Every: 

“I am not suggesting that we are guaranteed[?] to resolve every moral controversy through science. “ 

“Having received tens of thousands of letters and emails from people at every point on the continuum between faith and doubt, I can say with some confidence that a shared belief…” 

Undeniable: 

“Some version of this progression [of evolutionary morality] has occurred in our case, and each step represents an undeniable enhancement of our personal and collective well-being.” 

“It is undeniable, however, that if one side in this [9/11 conspiracy] debate is right about what actually happened on September 11, 2001, the other side must be absolutely wrong.”

I've put question-marks after words that are near a negation, because those might have been picked up by the filter and not taken into consideration. So, at least, the caveats (“I’m certainly not claiming”, “I am not suggesting that we are guaranteed”), might have been filtered out and not worked “against” the writer. Still, it's pretty clear that language is so flexible that “certainty” can be used to many different effects, that have little to do with the kind of “certainty” Haidt wants to “accuse” “New atheists” of.

As the commentator says, there is one example here that seems pretty “dogmatic”. I'd still argue it's sound reasoning, but even if I concede that, there’s so much “noise” taken as “evidence” that I’m not sure how accurate this analysis could be in showing that Harris exhibits more “certainty” than others. Isn’t all of this awfully arbitrary?

Then, as another commentator points out:

“The word “fundamentalis*” is included in the list. Much of End of Faith is about religious fundamentalism! How is this word analysis even remotely appropriate for gauging certainty on the author’s part? A proper word analysis would have to take into account the context of each word on the list (a lot more work) and then be categorized into 2 buckets: 1.) words that display certainty on the author’s part 2.) everything else.”

“The End of Faith” is sure to contain that word countless times, with very little relation to the author's “certainty” about anything. I could write a book about fundamentalism that is entirely descriptive, and I might score higher than anyone. Or, better yet, I could write a book about relativism and go on and on about how “New Atheists are more certain than they should be”, “they always talk in absolutes” etc and, again, I might score higher than anyone, religious, atheist, realist, and relativist alike.

As yet another commentator says:

“Let’s assume that `certainty words’ do measure what you argue they measure. Let’s take a hypothetical situation in which I write a book supporting a particular point of view, and X% of my words are found to be `certainty words.’ Let’s imagine I hire an editor, who takes out unnecessary, extraneous wording, cleans things up, makes things more precise, but does not change the number of `certainty words’ per argument. The edited book will have a lower X%, just by virtue of having a different writing style. Discuss”

But, I could even concede that, as well. Let’s assume the word count measures what Haidt claims it measures, let’s forget about writing style, let’s forget everything that’s been said.

Thirdly, and maybe even more importantly, the content of the discussion is ignored. Even when you, personally, are claiming certainty, it doesn’t matter what you’re claiming certainty about. You’re just as wrong to claim certainty whatever it is you’re talking about; whatever the scientific level of certainty for that claim. And that is, essentially, the enormous misconception that runs around by the name of relativism.

The “New Atheists”, and most scientists that speak against unfounded faith-based truth claims, always offer the caveat right away: “There is never 100% certainty”. But, also, scientists understand that, by default, the higher the probability, the higher the certainty of the claim. So an explanation that is 99% probable cannot be treated as equal to another which is 1% probable.

On the other hand, most intelligent people who energetically support faith-based claims are very accommodating. Their entire argument is that “I don't know for sure, you don't know for sure, so our positions are equally valid and probable”.

So, when the subject in discussion is the age of the Earth, the theory of evolution... then the relativists are the ones who say “I'm not claiming I'm sure! Both opinions are equally probable!” and it's the realists that say “Actually, we've answered those questions. We're as sure as we can be that the Earth is not flat and we can now stop talking about it as if it's still under consideration.”

Let’s consider the following example of a dialogue:

“Water is a chemical compound with the formula of H20.”

“Water is unicorn tears.”

“There aren't unicorns.”

“Do you know that for certain?”

“Yes.”

“Are you 100% sure? There’s no way you could be wrong?”

“Yes, unicorns do not exist.”

“Aren’t you being dogmatic?”

“Well…”

“Don't you always say that you can't ever be absolutely sure about anything? Aren’t you just as bad as I am, since you’re not willing to entertain the idea that you might be wrong? Science has been wrong before, you know! Are you absolutely sure that unicorns will never, ever be proven to be real?”

“No. You're right; it's equally probable that unicorns exist, as it is that they do not exist. And water might very well be unicorn tears. I have like tons and tons of scientific evidence from years and years of scientific research that clearly contradicts that, but I wouldn't want to be “dogmatic”, “close minded”, and a “hypocrite”, so I concede. Both explanations are on equal footing.”

Teach the controversy!

You can't be sure anyway!

Having to accept all ideas about everything as equally valid and/or probable, otherwise being called a “hypocrite”? This is a ridiculous standard to hold scientists against. I could even go so far as to argue that when we have total nonsense on one hand and scientific evidence on the other, it’s not even wrong to say you’re “certain”. But, even if we should ban all “certainty words” to appease accommodationists (a straw-man, I know), even if it was absolutely ALWAYS wrong to say you are certain (please note the inherent contradiction in this sentence)… clearly, one cannot be just as wrong in saying “I'm sure there aren't unicorns” as in saying “I'm sure there are unicorns”. The former is leaving out a 0.00000000000000000001% probability (if even that) that there are, indeed, unicorns, while the latter is standing against 99.99999999999999999999% of probability that there aren't.

If someone has trouble grasping this concept I advise them to take a course in statistics; and, for their sake, to never gamble.

I shouldn't be "angry"?
You can't derive an "ought" from an "is"!

The same goes for the expression of emotions. Just the use of the tired stereotype of “New atheists often sound angry”, is probably enough to expose Haidt's own biases. Even if I conceded Haidt's “impression” that this is indeed anger (I would call it urgency)... it's a very common misconception that you’re just as wrong in “showing any and all emotions”, whatever the situation.

Now, it's true that if most your arguments are appeals to emotion, I will not be taking you very seriously. However, Dawkin's and Harris' books aren't scientific papers, and there is something to be said about the correspondence of situations and emotions (see the psychiatric concept of “appropriateness of emotion”).  One is angry because you’re not allowing them to rape you, you’re angry because they want to rape you. One person is angry because we don’t teach our kids creationism as a real explanation, another is angry because some people want to teach creationism as a real explanation; tomato, tomahto. If ignorance and selfishness don’t manage to be the end of us, relativism will happily do the job.

Dear me, lower your voice, you're being impolite and annoying.

In conclusion, the only thing this word count shows for sure is how much one talks about certainty (and uncertainty), regardless of how certain (and uncertain) they themselves are. And that makes it wholly unsurprising and irrelevant that people like Richard Dawkins and Sam Harris score high on that account. It’s pretty much like analyzing cookbooks and concluding that they talk a lot about food.

So, Haidt’s essay could be summed up like this: Sloppy at best, intentionally misleading at worst.

But I'm sure my essay comes off angry, spiteful, and hating. So, I did my own analysis and compared it to Haidt's: I scored 4.22 in positive emotions, while Haidt only scored 3.61. That clearly proves that I'm a happier, kinder person. There.

I only have one more question... how can anyone be a relativist without considering even relativism… relative?

Isn’t it equally probable that you’ve got it all wrong?

Don't you ever worry you might only be holding on to your own precious preconceived notions and hindering real, tangible progress in the meantime?

Tatiana Valeonti
Med Student