Last Friday, the Irish Examiner published an opinion piece by pop psychologist Tony Humphreys, presenting a re-hash of the now anachronistic view that autism is caused by poor parenting (the so-called “refridgerator mom” theory). Oh, and despite claiming to explain the causes of autism, he simultaneously sought to argue that autism doesn’t exist.

Several authors have already described the various factual inaccuracies in the Humphreys article, the unwarranted distress it causes to parents of children with autism, its stigmatization of people based on their personality or occupation, its stigmatization of persons with needs for mental health and psychological services, its promotion of scientific myths, and its misrepresentation of the field of psychology (both professional and academic).

However, whatever about the errors in its subject matter (which are manifold), there is at least one sense in which the Humphreys article can be welcomed. For it gives us a wonderful opportunity to examine the various ways in which people in positions of professional authority can end up talking rubbish. In fact, seldom have I seen so many logical fallacies crammed into such a brief piece. It really is an excellent teaching tool, and I highly recommend it to teachers of critical reasoning, logic, epistemology, clinical decision-making, and scientific communication.

Therefore, I present to you: Tony’s Ten Top Tips for mounting illogical arguments…

* * *

#1. Imply that ‘facts’ are arbitrary

Here is Humphreys expressing shock at some research by Cambridge neuroscientist Simon Baron-Cohen:

What is shocking is that Dr Baron-Cohen and the team of researchers are… assuming that autism is a scientific fact…

“Assuming that autism is a scientific fact”…is a strange construction, in that it equates a noun (autism) with a truth-statement (fact). I guess what is intended here is a claim that the team are “assuming that the existence of autism is a scientific fact”. Even so, this is still strange. Why not simply say “assuming that autism exists”? After all, that is what he is getting at.

The term “scientific fact” is a misnomer, because the word “fact” is unqualifiable. There aren’t different ‘kinds’ of fact. Either something is a fact (such as an axiomatic truth) or it isn’t a fact (such as a theory, hypothesis, belief, or empirically supported position). Terms like “scientific fact” are often used by critics of science to imply that what scientists have to say cannot really be trusted. However, the term is incoherent, and its use is a sign of an unfamiliarity with how scientific research works.

* * *

#2. Appeal to common belief (or employ argumentum ad populum)

Humphreys continues to express shock at the Cambridge researchers as follows:

…[Dr Baron-Cohen and the team are] missing the glaringly obvious fact that if the adults they researched live predominantly in their heads and possess few or no heart qualities, their children will need to find some way of defending themselves against the absence of expressed love and affection and emotional receptivity.

By declaring this to be “glaringly obvious”, Humphreys is asserting that we need not look for evidence to test whether or not it is in fact true. Any claim that begins with the implication “it is glaringly obvious…” is unlikely to be scientifically sound, or even – in plainer terms – to be logically coherent. If something has to be signposted as “glaringly” obvious, then it almost certainly is not that obvious at all, or at least will not be obvious to some people. Such phrasing indicates a type of argumentation described in epistemology as an appeal to common belief. It is an unsavoury and almost fascistic type of argument, because it warns the listener in no uncertain terms that contrary views will not be tolerated. It is purely rhetorical and scientifically nonsensical.

* * *

#3. Present opinions as though they were facts

Here is a key passage in Humphreys’ overall argument. It is critical because its content represents his overall justification. If this passage is not true, then his entire argument falls:

Children’s wellbeing mostly depends on emotional security – a daily diet of nurture, love, affection, patience, warmth, tenderness, kindness and calm responses to their expressed welfare and emergency feelings. To say that these children have a genetic and/or neurobiological disorder called autism or ASD (autistic spectrum disorder) only adds further to their misery and condemns them to a relationship history where their every thought and action is interpreted as arising from their autism.

However, when Humphreys states that “Children’s wellbeing mostly depends on emotional security” and that a diagnosis of ASD “only adds further to their misery”, he is presenting opinions rather than facts.

It is entirely reasonable for a person to earnestly argue the opposing positions, namely, that (a) children’s wellbeing depends mostly on other factors, and (b) a diagnosis of ASD actually reduces misery. Therefore, in order to sustain his argument, Humphreys needs to corroborate his statements. However, he doesn’t. In fact, he offers these opinions as though they were facts. This is a hallmark of pseudoscience.

Moreover, in professional contexts, it is unethical. The code of ethics of the Psychological Society of Ireland, the professional body for Humphreys and his psychologist colleagues in Ireland, states that when communicating knowledge professional psychologists must “clearly differentiate facts, opinions, theories, hypotheses, and ideas.” Humphreys appears not to do this.

* * *

#4. Present claims that cannot be falsified (or, try to have it both ways) 

Here Humphreys describes the processes by which he says parents and teachers end up damaging the children they are caring for:

An unconscious collusion can emerge between parents and teachers to have these children psychiatrically assessed… the childrens’ lives are not examined and their mature development is often sacrificed on the fires of the unresolved emotional defences of those adults who are responsible for their care.

Terms like “unconscious collusion” and “unresolved emotional defences” refer to a view of psychology presented in psychoanalytic theory. Psychoanalytic theory is considered logically problematic (to say the very least) precisely because it invokes the idea of unconscious forces. Because such forces are, by definition, unobservable, it is impossible to establish whether or not they exist. As a result, it is impossible to falsify a psychoanalytic hypothesis. Therefore, it is impossible for a psychoanalytic theory to be coherent. So Humphreys’s position is incoherent.

Many psychoanalysts feel that such critiques represent unfair attacks of their ideas by (perplexed) scientists. However, this criticism doesn’t come from scientists; it comes from philosophers, and in particular from those concerned with logical reasoning and epistemology. Scientists, to be blunt, don’t give fields like psychoanalysis a second thought.

* * *

#5. Present anecdotes rather than evidence (or, make hasty generalizations) 

This is how Humphreys attempts to convince readers that he is definitely right:

Indeed, my experience in my own psychological practice is that when parents and teachers resolve their own fears and insecurities, children begin to express what they dare not express before their guardians resolved their own emotional turmoil.

Rather than point to empirical evidence from disinterested observers which might corroborate his position, Humphreys instead refers to “my experience in my own psychological practice”. This is a clear valorization of anecdotal evidence, which, in philosophy, is normally discussed as an example of fallacious logic.

In some ways, it is perfectly obvious why anecdotal evidence is unreliable. For one thing, who exactly frequents Tony Humphreys’ “psychological practice”? I doubt that the people doing so constitute a nationally representative sample of parents and children. Therefore, even if Humphreys were an unusually astute observer of his own clients, it should be obvious even to him that his capacity to generalize from these observations must be fatally limited. (And if he’s not astute enough to realise this, then why should we trust his anecdotes?)

Anecdotal positions are also subject to several basic errors of remembering and interpretation on the part of the person telling the anecdote. In summary, ignoring its obvious limitations suggests that the writer either doesn’t know or doesn’t care that anecdotal reasoning is intrinsically unreliable.

* * *

#7. Commit category mistakes

Ironically enough, when attempting to correct others on matters of fact, Humphreys say this:

A clear distinction needs to be made between the autism described by psychiatrist Leo Kanner in 1943 and the much more recently described ASD (autistic spectrum disorder, often referred to as Asperger’s syndrome).

Humphreys says that ASD is “often referred to as Asperger’s syndrome”. It isn’t. This is not just a demonstrable error of fact, it also reveals a type of flawed reasoning. Specifically, it is a type of error known as a category mistake. Asserting that ASD is referred to as Asperger’s syndrome is akin to saying that vegetables are “often referred to as apples” or that animals are “often referred to as tigers“. If a child said these things, you would no doubt offer a correction. But you would probably do so sympathetically, understanding that she or he is still just developing a proper grasp of the relevant grammatical structures.

However, Humphreys is not a child. In his case, the category error reveals nothing less than a startling unfamiliarity with the subject matter he has chosen to write about.

As does the following…

* * *

#8. Appeal to dualism

Humphreys presents the following observation:

Curiously – and not at all explained by those health and educational professionals who believe that autism and ASD are genetic and/or neurobiological disorders – is the gender bias of being more diagnosed in boys (a ratio of four to one).

This sentence is indeed curious. Embedded within is the following argument: “A gender bias in diagnosis is not (at all) consistent with the belief that autism and ASD are genetic and/or neurobiological.” In essence, Humphreys is stating that if something differs by gender, then it cannot be genetic and/or neurobiological.

This is truly bizarre. It would be more accurate to state the opposite: if something differs by gender, then it is quite likely to be genetic and/or neurobiological. Here are some examples: facial hair, height and build, presence of sex organs, susceptibility to cardiovascular disease, etc.

Arguing that if something is a psychological difference then it cannot also be a biological difference is an example of an appeal to dualism. As a purported expert, for Humphreys to adopt such an argument betrays a command of his subject that is deficient on a truly breathtaking scale.

* * *

#9. Use absolutist statements

In complaining about diagnostics, Humphreys states that there is:

[an] absence of any scientific basis or test for either the originally ‘detected’ autism or for the broader construct of ASD.

By stating that there is not “any” scientific basis to such things, Humphreys is presenting a type of claim known as an absolutism. Absolutist statements tend to be avoided in logical discourse, because they are ultimately unreasonable. Just one exception will render your claim false; therefore, you should never use an absolutist statement unless you have good reason to believe that no such exception exists. There are in fact a number of scientific bases for distinguishing the diagnostic categories of autism and ASD. There are also several tests. Therefore, Humphreys’ statement is not just poorly constructed scientifically, it is also poorly informed factually.

* * *

#10. Misrepresent your sources

Humphreys cites a book by Timimi and colleagues in support of his overall position:

What they are saying is – and I concur with them – that the focus needs to be on the relationship contexts of these children’s lives, and to take each child for the individual he or she is and to examine closely the family and community narratives and discover creative possibilities for change and for more dynamic and hopeful stories to emerge for both the children and their carers.

Humphreys is arguing that Timimi’s claim that “the focus needs to be on the relationship contexts of these children’s lives” and on their “family and community narratives” corroborates his theory that autism is a condition caused by poor parenting. But the Timimi book does not support this at all. In fact, this is what the authors assert regarding such theories (p. 293):

We are NOT saying any of the following: 1. Autism is a condition caused by poor parenting…

How Humphreys interprets the conclusion “We are NOT saying that autism is a condition caused by poor parenting” as corroboration for his theory that “Autism IS a condition caused by poor parenting” is unclear.

* * *

So there you have them. Tony’s Ten Top Tips for mounting illogical arguments. (Creating unnecessary distress in vulnerable groups while stigmatizing families living with autism is purely optional.)

In response to what has been a wave of public complaints, Monday’s Irish Examiner featured an editorial asserting that people like Tony Humphreys have an entitlement to freedom of expression. I actually agree with that position. The Irish Examiner can indeed publish what they choose.

Of course, it is interesting to note that they are currently choosing to publish at least some articles that are so incoherent and illogical as to stand as actual case-studies in bad argumentation.

At least that’s a contribution I suppose…

=  =  =  =  =  =  =  =  =  =  =  =  =

Post-script, 9 Feb 2012: Congratulations to the reader who spotted that my so-called “Ten Tips” doesn’t contain a number 6! The only explanation I can offer is that my brain was addled by all the mangled logic I was trying to summarize. Suffice to say, there are several possibilities for that tenth slot, which space prevented me from including. These include appeal to authority, straw man argumentation, cherry picking, moving the goalposts, single-cause fallacy, etc., etc. Some other suggestions appear in the comments below. I’ll do better next time!