Wednesday, July 27, 2016

What is and isn't Ad Hominem

Perhaps the most common logical fallacy, Ad Hominem is an attack on the source of an argument instead of analysis of their argument. It's a fallacy because a bad source may have a good argument and vice versa.


There are many cases which look like ad hominem, because the source of the argument is part of the analysis of the argument, but aren't ad hominem. For example:

YouTuber Thunderf00t, after making a video criticizing Hyperloop, gets into a Twitter fight with (from what I gather) a SpaceX employee:

(I gathered incorrectly; Casey Handmer is a Hyperloop One employee. In my defense, his profile header is a diagram of a rocket engine... I'll add corrections where appropriate below, in blue.)

At this point one might ask the --- apparently ad hominem --- leading question
Does Thunderf00t believe that he, who is not an engineer, noticed "a lot of" engineering problems that a number of SpaceX Hyperloop One engineers didn't? 
(According to the preamble to interaction above, SpaceX Hyperloop One uses the Hyperloop Alpha paper as the basis for a recruiting question; also, since it's a pet project of Tony Stark Elon Musk, many SpaceX engineers will try to impress the boss with Hyperloop-related ideas or comments. Yes, I'm leaving that last SpaceX unchanged because what I meant was that engineers working on the rocket side at SpaceX would try to impress their boss by showing initiative on a different project.)

This question has some of the hallmarks of ad hominem: it appears to accuse Thunderf00t of putative arrogance and therefore appears to be an attempt to discredit his arguments without addressing them. And it would be ad hominem if that was the end of the process.


It's clear, a few minutes into the video that Casey Handmer is responding to, that Thunderf00t doesn't know much about engineering or operations management:

(For example, and just off the top of my head, he seems to think that the track will be a 600km tube with no isolatable zones, no shunt lines, no emergency, safety, or maintenance affordances; in other words, nothing like a railroad, which is what Hyperloop is.*)

Ad hominem is attacking the source in lieu of the argument. Noting that the source doesn't know the basics of the field and therefore estimating a low return from investing the time to follow an argument built without knowing the basics, isn't ad hominem. It's effective use of time and attention.

To be clear: this isn't ad hominem because the decision isn't based on {Thunderf00t isn't an engineer} but rather on a probabilistic determination of who is more likely to be wrong, using the new information (the video) to update prior probabilities based on the past observations: {Thunderf00t has correctly pointed out problems with other "innovations"} and {SpaceX does a lot of brilliant engineering} {Presumably not all engineers at Hyperloop One are completely incompetent}.

A  managerial critique of the video is that it treats the preliminary, idea-phase, design as if it was a final specification, whereas Hyperloop hasn't even finished the technology demonstrator phase of their innovation path. Usually this kind of error is made by people investing in the technology; ironically enough, Thunderf00t uses examples of that error to mock another twitterer.

I think that Thunderf00t's style of argumentation lends itself to these Dunning-Kruger traps. (Common to many e-atheists and science popularizers.)  Destin "Smarter Every Day" Sandlin makes great science and technology popularization videos by taking the opposite tack:
(1) Destin doesn't assume that his personal knowledge base is enough to do everything himself: he asks others, experts in the fields, to explain stuff that he mostly already knows, for the benefit of the audience. 
(2) Destin comes across as a genuinely nice person whose videos focus on the science and technology, instead of snark and mockery which may work with Thunderf00t's core audience but fail to convince grown-ups.
Here's how Destin responds to a massive pig's breakfast of a statement by a popular science popularizer:

Nice guy attitude all the way through. All  about the science. Brings in a subject matter expert, despite clearly knowing the material himself. Includes the response of the person being corrected and thanks him. Class act, no snark.

Then again, to some audiences, snark sells. Just doesn't sell actual knowledge.

-- -- -- --

* There are many practical issues with the Hyperloop concept, some of which may be dispositive (against it), but none of these is mentioned in Thunderf00t's video. For example, I question the ability of Hyperloop to get the rights-of-way needed for construction, given how anti-change California municipalities are (unless you have pull, of course).