Affirm the Consequent, Deny the Antecedent, and don’t mess with the Excluded Middle.
November 5th, 2007
A lot of people think that Logic is for Debate Team nerds and people who like to torture undergraduates on warm Spring afternoons.
It is all that, and more!
When I was forced to take Dr. Sorenson’s Logic 1A course, I was expecting to be bored to near the brink of death - I expected to see the long, dark tunnel, bright light and long-dead relatives just before the bell rang for the end of each class.
I was wrong.
What I learned was not how to score geek-points in Debate, but how to be alert to people trying to fool me. Dr. Sorenson taught us using clippings from the daily newspaper, showing us how advertisers, political campaigns and even reporters and editors were using logical fallacies to sound convincing while deceiving us.
In the past, we have covered some of the more egregious fallacies (and the most common), particularly the Post hoc, ergo propter hoc (”After this, therefore because of this.”) fallacy and its close cousin, Cum hoc, ergo propter hoc (”With this, therefore because of this.”) which together are fallacies of false causation (Non causa pro causa - “Non-cause for a cause”).
The fallacies I’d like to discuss today are a bit harder to grasp, perhaps, than false causation, but they are just as ubiquitous in the world, including the world of “alternative” autism therapies.
Affirming the Consequent / Denying the Antecedent:
These two fallacies are two complementary halves that make up a lot of the nonsense spouted about autism and its so-called “treatments”.
For example, there has been a lot of talk about how certain patterns of urinary porphyrins are “diagnostic” of mercury toxicity. This is true to the extent that James Woods of the University of Washington has published in a number of articles on how mercury affects the excretion of urinary porphyrins (here, here and here). Of interest, he has also recently published an article showing that mercury from dental amalgams has no sigificant impact on neurobehavioral measures in children.
A number of people have claimed (and at least one laboratory is selling the tests) that this “pattern” of urinary porphyrin excretion is not only consistently found in mercury exposure (as Dr. Woods has convincingly shown), but that it is diagnostic of mercury exposure.
This is where the fallacy lesson comes in.
What Dr. Woods has shown (and, I believe, all that he is claiming) is that the pattern of urinary porphyrins is seen consistently in a setting of mercury exposure. This is not the same as saying that this pattern indicates mercury exposure.
Let me illustrate:
When my face gets sunburned, it turns red.
Therefore: If my face is red, it must be sunburned.
If I say that my face gets red whenever it gets sunburned (a consequence of my Northern European heritage), it does not mean that when my face is red, that it is always sunburned. I might be blushing, or hanging upside down, or exerting myself.
This is affirming the consequent. Just because “B” happens everytime “A” occurs (B = face getting red, A = sunburn) does not mean that whenever “B” (red face) happens that “A” (sunburn) has occurred.
In the case of the urinary porphyrins, the setup for the fallacy is something like this:
Mercury exposure causes a charateristic pattern of porphyrin excretion.
Therefore: If a person has the characteristic pattern of porphyrin excretion, they must be mercury toxic.
As it turns out, there are a lot of possible reasons why someone would have the “chracteristic pattern” of porphyrin excretion, including (but not limited to), other toxic metals, other toxic exposures and genetic abnormalities involving the enzymes engaged in porphyrin synthesis and metabolism.
This is not just idle speculation. One of the enzymes that is affected by mercury, causing the “characteristic pattern of porphyrin excretion” is uroporphyrinogen decarboxylase (UROD). UROD just happens to be the enzyme involved (impaired, as it turns out), in porphyria cutanea tarda, the most common of the porphyrias (see here).
The other enzyme Dr. Woods has implicated in the “characteristic pattern of porphyrin excretion” is coproporphyrinogen oxidase (CPO). This enzyme is impaired in heriditary coproporphyria (see here).
So, we have known disorders that involve impairment of the two enzymes that are responsible for producing the “characteristic pattern of porphyrin excretion”. And how does mercury cause this “characteristic pattern of porphyrin excretion”? By imparing the two enzymes mentioned above.
So, it is not just hypothetical that a genetic abnormality - a mutation - could cause the same “characteristic pattern of porphyrin excretion” as mercury.
To the best of my knowledge, Dr. Woods is not claiming that this “characteristic pattern of porphyrin excretion” can be used to diagnose “mercury poisoning” when there is no significant exposure to mercury. I suspect that others have mis-interpreted his findings to their own ends.
The “flip side” to affirming the consequent is denying the antecedent. Let me illustrate:
I only get headaches when it rains.
Therefore: If I don’t have a headache, it must not be raining
If I say that I only get headaches when it rains, that does not mean that if I don’t have a headache that it isn’t raining.
The Excluded Middle:
More properly called the false dilemma, this fallacy has been popular with politicians (”You’re either with us, or you’re against us!”) and demagogues since before recorded history. It consists of simply presenting two distinct choices and asserting that they are the only choices available.
I commonly run across this fallacy when parents claim that I am accusing them of lying when I doubt their stories (see here). What they have done is excluded the middle choices. In this particular example, the choices II am presented with are:
[1] The parents are presenting an accurate description of the events.
[2] The parents are deliberately telling a false description of the event (lying).
In fact, they have omitted at least one middle choice:
[3] The parents are telling what they believe to be the truth, which is an amalgam of their observations and their interpretations of these observations, their observations having been edited by their memory of the event, events preceding the events in question and events which have occured subsequently.
The same sort of fallacy arises in many of the testimonials that “alternative” autism therapy uses in place of data. The usual setup is something like:
“If you don’t treat your child with [fill in the blank] they will end up [a] in an institution [b] in a fate worse than death [c] in a life not worth living.”
These people have omitted the middle options of (not an exhaustive list):
[d] having a fairly normal life.
[e] having a happy and productive life.
[f] running a large multi-billion dollar software company.
[g] getting better anyway.
[h] being the best human being they can be.
[i] being happier than you are.
[j] …..
Prometheus
Filed under: Autism Science, Critical Thinking
November 5th, 2007 at 11:51 pm
Excellent logic examples! It would be really cool if they taught this in high school (perhaps they do now).
November 6th, 2007 at 4:13 am
Another master class from Prometheus. Thank you.It is a pity that more of the mothers posting on YaWoo groups for help interpreting their kids’ porphyrin test results do not read this blog.
November 6th, 2007 at 1:37 pm
Excellent examples for both fallacies. I hope some of the false dichotomy reader/responders take the logic to heart.
November 6th, 2007 at 4:02 pm
Hi Promotheus -
Good stuff w/the logic.
“So, it is not just hypothetical that a genetic abnormality - a mutation - could cause the same “characteristic pattern of porphyrin excretion” as mercury.”
Even though many times symtpoms are not noted; shouldn’t some people with abnormal porphryin profiles that are the result of inherited or aquired PCT be showing signs the disease, such as “Areas of skin exposed to the sun develop severe blistering, scarring, changes in pigmentation, and increased hair growth.” (via Wikipedia?) The population we are interested in certainly doesn’t seem to exhibit these symptoms at a greater rate than normal from what I’ve seen or read. (?) [It sure isn't in the DSM!]
As I understood it, (?) people have seen profiles change after being chelated. To my mind, this was the most striking thing about the infamous Nataf paper; not just that abnormal profiles were found in autistics compared to controls, which could be from a variety of factors, but that the profiles were modified after treatment tailored for removing heavy metals.
Unless we have a mechanism by which chelation could be affecting porphyria cutanea tarda or heriditary coproporphyria, doesn’t this indicate the profiles (in these cases) were the result of metal exposure, or something else, but almost certainly not genetic problems with the enzymes noted?
I read a bit, and it seems that Porphyria Cutanea Tarda can also be caused by exposure to some agents like alcohol, an excess of iron, as well as viral infections. Perhaps (?) the removal of iron from chelation, even by agents not tailored for it, and “not mercury” is the cause of a change in profiles, though this would beg the question as to why one population would have more iron than another. While DMSA does appear to be able to remove iron from the blood, several studies I found on Iron and autism in pubmed seemed to indicate a relative increase in iron deficiency as compared to the other way around. These were serum studies; perhaps for whatever reason, the amounts in the liver, which is what tends to be increased in PCT are greater. (?)
I’m trying to show that I’m open minded, but at some point, don’t we have to question how many maybes we must exhaust before we go back to the simplest explanation?
Exposure to mercury, among other things, have been shown to result in abnormal profiles. Appplication of agents designed to remove mercury have been shown to normalize profiles. Considering the frequency of PCT (1 in 25K), and the lack of shared physiological effects of those diagnosed with PCT and autism, why should we consider that genetic, or aquired Porphyria Cutanea Tarda or coproporphyria is the cause of abnormal porphyrin profiles in autism? Because it might be? This is the kind of reasoning you usually attack.
““If you don’t treat your child with [fill in the blank] they will end up [a] in an institution [b] in a fate worse than death [c] in a life not worth living.””
I get a ton of emails from list servers, read a few of them, but have never, ever seen anything like this. I seem to live in a bubble of extreme coincidence, however, so perhaps this is yet another chance occurrence.
btw - The link to coproporphyria on photoninthedarkness.com gives me a 404.
Take care!
- pD
November 6th, 2007 at 5:13 pm
Thank you Prometheus. This post came just at the right time as I have stirred up trouble on an advocacy listserve I belong to. Someone made reference to a Geier article and I dared (gasp!) question the ‘researchers.’ I smoked out the fallacy followers faster than you can say, “porphyrin excretion, say wha..?”
I loathed “Fundamentals of Reasoning 101″ my freshman year in college, an elective. I got a C-, the lowest grade of my undergrad career. How I wish I had paid better attention OR waited to take it while in grad school.
November 6th, 2007 at 5:48 pm
pD,
You are making the unwarranted assumption that all mutations of these genes would have the same manifestation as PCT and hereditary coproporphyria. In fact, a number of studies (including some by Dr. Woods) have shown that there are variations in these genes that lead to smaller changes in their efficiency without overt symptoms.
As to the effect of chelation on porphyrin profiles, I would like to see those results replicated by someone other than the Geiers. Until then, we’ll have to take their findings with a grain of salt.
It’s really quite simple and doesn’t require any “maybe’s” - there are people (including the Geiers) who claim that these porphyrin profiles are diagnostic of mercury poisoning - in the absence of other indicators of mercury toxicity (e.g. urinary, blood or hair mercury levels). In fact, a number of authors have reported that autistic children have lower mercury levels than neurotypical controls.
So, until there are data showing that this porphyrin profile is specific for mercury, it is premature to use it as a “test” for mercury toxicity.
I’m not saying that it’s not possible, just not proven.
If you haven’t seen the “If you don’t treat your child with [blank], they’ll end up…”, line yet, then you haven’t been looking. Check with John Best - he’s said it a bunch of times (and worse than that). I’ve seen it hundreds of times over the years.
Also, thanks for the “heads up” on the link - it should be fixed now.
Mekei,
You’re welcome. From what I’ve heard, Professor Sorenson was an exception to the rule of Logic professors. I was lucky.
Good luck in you efforts to spread a little healthy skepticism!
Prometheus
November 7th, 2007 at 9:46 am
Thank you for these clear explanations. I’m learning a lot through your excellent posts.
I’ve downloaded a whole list of logical fallacies to go through.
It’s driving my husband mad though; whenever we’re having a ‘rigorous discussion’ I point out when he uses these fallacies! He he!
November 7th, 2007 at 10:28 am
Hi Prometheus -
“You are making the unwarranted assumption that all mutations of these genes would have the same manifestation as PCT and hereditary coproporphyria. In fact, a number of studies (including some by Dr. Woods) have shown that there are variations in these genes that lead to smaller changes in their efficiency without overt symptoms.”
Hm. Point taken. Considering I personally know at least four children in who have the abnormal profile, all of which have autism, this would tend to point in the direction of this mutation having a relationship to autism. (I don’t work in the autism field, but rather know many children from fellow parents).
I guess I felt you were leaving out a critical point; i.e., If I have a red face, and it feels better when I apply lanolin, then it might be a sunburn. I have no particular interest in defending the ‘must be’s’ of the Geiers or anyone else. As often seems to be the case, I seem to be arguing for a more nuanced approach, while you seem (?) to be finding the most extreme viewpoints as foils.
“If you haven’t seen the “If you don’t treat your child with [blank], they’ll end up…”, line yet, then you haven’t been looking. Check with John Best - he’s said it a bunch of times (and worse than that). I’ve seen it hundreds of times over the years.”
Pointing out the logical flaws in arguments made by Mr. Best is beneath you, Prometheus.
Take care!
- pD
November 7th, 2007 at 1:40 pm
Sharon,
Be sure to check out the Fallacist’s Fallacy (http://www.fallacyfiles.org/fallfall.html). It points out that the strength of the argument - whether or not it contains logical fallacies - does not affect the truth of the conclusion. In other words, people can use bad arguments to support conclusions that are true.
pD,
I haven’t concluded that the urinary porhpyrin tests don’t reliably diagnose mercury toxicity, only that they haven’t yet been shown to. There is still some work - a lot of work - needed to determine their false positive rate.
As for chelation being a diagnostic tool, even Vas Aposhian, who has done more work with DMSA and DMPS than any single person I can think of, has argued that these compounds have many known and potential effects. It is foolish to presume that any improvement seen after their administration is solely due to their metal chelating action.
For that matter, it remains to be seen if the improvements seen after chelation are related to the chelation. A recent presentation by Jim Adams showed no significant difference between one and seven “rounds” of chelation, yet many parents report continuing improvement after years of chelation.
This needs to be studied further before any conclusions can be made. Remember that practitioners and parents were claiming that 70% (and more!) of children were “significantly improved” by secretin - which we later found was no better than placebo, even in a highly selected group of children.
I’ll agree that John Best is beneath me, but he is not the only person who has made that assertion - just the only one whose name I remember. Look around - you’ll find that argument used by a number of people.
Prometheus
November 8th, 2007 at 1:31 pm
Assuming it is confirmed that chelation lowers porphyrin levels in autistics, the next question at that point is whether it lowers them in the general population as well. Chelation moves a lot of stuff around. Metabolic changes are expected, and the mechanism may not necessarily be what it has been speculated to be.
November 9th, 2007 at 1:38 am
It’s not just viral infections which can produce porphyria; infections with the bacterium Chlamydia pneumoniae have also been noted to do so, presumably by stealing enough host ATP to interfere with heme production, thus causing porphyrins to build up. C. pneumoniae is suspected of causing both multiple sclerosis — a disease of adults and of older children — and Alzheimer’s disease — a disease of the elderly. An obvious question, in light of those hypotheses, is what disease it might cause in young children, when it infects their central nervous systems.
Intriguingly, DMSA, one of the main agents used in chelation therapy, has been found to be highly active against the ‘elementary body’ (EB) stage of the life cycle of C. pneumoniae. EBs are the spore-like forms in which the infection is spread from person to person (as well as within the body). Their outer wall consists of proteins held together by disulfide bonds; agents such as DMSA or N-acetyl-cysteine can reduce those bonds and break up the outer wall. So if C. pneumoniae were causing autism, it would be entirely understandable for chelation with DMSA to have a beneficial effect.
(This is not the sexually transmitted disease; that’s Chlamydia trachomatis. Chlamydia pneumoniae is a common cause of upper respiratory disease; as an obligate intracellular organism with a very slow growth rate, it escaped characterization until a couple of decades ago. Since then it has been implicated in a variety of slow-moving disorders, most notably atherosclerosis. It is very difficult to extirpate, as most antibiotics tend not to kill it but only force it into a quiescent state.)
November 9th, 2007 at 2:35 pm
Another excellent post.
December 22nd, 2008 at 4:09 am
[...] of rapid development followed by periods of stasis or even some regression. Some children–perhaps as high as 19% of them–even improve enough to lose their diagnosis; i.e., they improve to the point where they no [...]