Free Speech, Top Stories

“You’re Fake News” The Unfortunate Reality of the Ad Hominem

When Donald Trump tweets, stock prices can tumble. Trump can wield greater influence with 280 characters than some world leaders can with entire economies. Reaching the public directly, Trump is able to personally attack an individual, agency, or company, and impact the news cycle for days, if not weeks, at a time. How can Trump’s attacks be so effective if the restrictions imposed by Twitter’s character limit leave so little room to formulate an argument?

Anyone who followed the bitter presidential race between Trump and Hillary Clinton in 2016 could be excused for thinking the evidence is already in. Trump and Fox spent months ceaselessly browbeating CNN and Clinton with attacks on their integrity, their associations, and their alleged motivations. Attacks ranged from accusations of corruption and criminality to anti-American intent. Clinton and CNN responded in kind, drawing parallels between Trump and Hitler, and painting him as exploitative and predatory in his business practices, narcissistic, sexist and racist.

Behind these shouting matches, various news channels were busily scrutinizing every statistic, accusation, and proposal, consulting experts and available data to ‘fact check’ these accusations on behalf of the American people. But, despite these attempts at accuracy, Americans’ trust in the mass media “to report the news fully, accurately and fairly” has reached its lowest level in polling history, with only 32 percent of respondents claiming they have “a great deal” or “a fair amount” of trust in the mass media. It appears that, armed with ad hominems and an insatiable blood lust, news organisations on all sides have engaged in all-out partisan warfare which has cost them dearly.

An ad hominem attack is one directed at an opponent’s character rather than the substance of his arguments and claims. In a recent PloS-one article, researchers Barnes et al. investigated the effects that ad hominem attacks have on people’s attitudes toward scientific claims. The researchers presented college students and adults with claims that either 1) attack a claim directly, 2) attack the researcher (e.g. relevant misconduct, past misconduct, or conflict of interest), or 3) attack both the data and the researcher.

Example of the effect of ad hominem attacks on the evaluation of claims (Barnes et al., 2018)

Their results indicate that attacks on an individual’s credibility as a reliable source may be just as effective as attacking the claim that the person is making. Furthermore, combining ad hominem attacks with empirical counter-arguments seems to be no more effective than attacking the empirical claim in isolation. The results were the same for both college students and adults, suggesting that this effect may extend beyond the college campus and into the US population at large.

What’s striking about these results is their implication regarding the disturbing trends currently plaguing our society. Rewind to 2016, and the ad hominem arguments begin to make sense, not just from an academic perspective, but from a broader societal perspective. Trump attacked Clinton for being corrupt (i.e. misconduct), and at one point even suggested that if he became president, she would be thrown into jail. According to the research, the damning implication that she had engaged in criminal behavior worthy of prison time could be just as effective as actual evidence of criminality, and no less influential. Trump also frequently attacked CNN and the Clintons, for putting the interests of their friends and associates before those of ‘hard-working Americans,’ which implied a conflict of interest. Clinton took the same approach, attacking Trump for exploitative business ventures and for putting his own selfish desires ahead of America’s national interests.

The authors found that adding ad hominem or empirical criticism to the mix once one had already been used did nothing to enhance the credibility of the existing claim. So then why bother going after the facts once you have attacked a person’s integrity to similar effect? These findings leave a bitter taste. First, they signal the effectiveness of lazy arguments and disincentivize more sophisticated discussion of important issues, including questions as important and consequential as who ought to be president of the most powerful country in the world. Second, they suggest that attempting to mitigate collateral damage after blood has been spilled has very little effect on voter perceptions.

This study carries serious implications for discourse beyond electoral politics, and into the sphere of academe and wider public discussion. Each month there appears to be a new story about a public intellectual who has had an event canceled due to alleged promotion of ‘hate speech,’ or who has been disciplined by a university for encouraging ‘unsafe’ debate, or who has been stigmatized as ‘alt right’ and consequently requires security just to deliver their remarks on campus. The list of recent examples could fill a separate article.

These rhetorical devices are designed to label someone as an undesirable; the bearer of untrustworthy ideas that cannot be debated at face value. They are ‘fake news,’ as Donald Trump would say. The findings of the Barnes et al. paper allow us to face an uncomfortable truth. If you can convince someone that Steven Pinker’s expressed concerns about political correctness are based on ulterior, sinister motives, or that Dave Rubin is an adherent of a narrow-minded and racist ideology, then you need not waste time checking whether the ideas they espouse have any merit or are backed by good evidence. This has several important implications.

First, if attacks on a person’s character are effective, and potentially irreversible even with the subsequent addition of facts, it becomes easy to discredit people wishing to tackle the difficult but important issues facing our society. This tactic has the effect of instilling fear about the reputational costs of speaking up, and discourages individuals from forwarding controversial or unpopular hypotheses. The result is that prevailing ideas remain unchallenged. Such attacks are now commonplace on both sides of the political spectrum, with journalists and academics categorizing intellectual opponents as Nazis or Communists as a way of stigmatizing an individual in the mind of a casual reader not personally familiar with their views.

Second, as Pinker has observed, relying on ad hominems surrenders a rhetorical cudgel to opponents that can be especially dangerous in the hands of far-Right groups. If ideas are left to fester unaddressed instead of debated openly, then individuals susceptible to those ideas may seize upon their “most toxic interpretations” and decide that those interpretations are in fact forbidden ‘truths’ which the establishment has conspired to keep hidden. Since the idea itself is not being disputed in real debate, a simplistic and disfigured version of that idea is allowed to develop, stripped of important context. These ideas then enter a cycle of vindication as others deny them, push back, and force nuanced discussion into an ‘us and them’ mentality, thus creating the rise of dangerous ideas to be used as blunt force instruments.

This leaves us with a harsh truth that we must nevertheless less take seriously. Lazy and abusive rhetoric is an effective means of silencing of healthy debate. As a result, polarization in Western Society is likely to get far worse before it gets better.


Dr. Elio Martino is a provisional psychologist, writer and podcaster, with a PhD investigating killing in combat. You can follow him on Twitter @SpartanMindPsy


  1. David Turnbull says

    Sorry, but several of these so-called ad hominem attacks in science are definitely relevant to the empirical claim. Science is built on trust. If a researcher has violated this trust in the past, there is every reason to discount any claim he/she makes. Very little science is replicated, so trust must be absolute. You cannot be a ‘mostly’ trustworthy scientist.

    • Um, no. Science does not work on Trust. It works on Scepticism.

      “If a researcher has violated this trust in the past, there is every reason to discount any claim he/she makes.”

      It doesn’t matter if scientist has got every single claim right in the past. There is no reason to ‘trust’ their newest claim.

      • David Turnbull says

        You misunderstand the necessity of trust in science. Experiments are not and cannot be replicated as a matter of course. Only totally unexpected findings are re-tested; there is just not enough manpower and resources. Therefore we trust that our fellow scientists are not misrepresenting or falsifying their findings.

      • Right Off says

        Reading Nomad has it absolutely right. The facts are the only thing, as any peer reviewer would (hopefully) tell you.

        • David Turnbull says

          I have been a peer reviewer (retired now). Scientists who lose trust rarely make it to the peer review stage. They lose funding and their research programs grind to a halt.

          I have been on granting committees. Once a scientist is known to be fraudulent, that’s pretty much the end for them. Competition for grants is ferocious and it’s a dog-eat-dog world.

    • Márton Vaitkus says

      I’m reminded of the Royal Society’s motto “Nullius in verba” – “On no one’s word”.

  2. “relying on ad hominems surrenders a rhetorical cudgel to opponents that can be especially dangerous in the hands of far-Right groups.” Replace this with the suggestion that the rhetorical cudgel can be especially dangerous in the hands of fringe groups regardless of politics and your intent will appear less biased and actually keep the focus on the disservice of ad hominem attacks to the public in general.

    • Less biased? 1) The far-right part was a Steven Pinker quote which is in context specifically referring to the far-right. I can’t simply put words into his mouth to appear more balanced. 2) I also discuss far left groups as a problem by labeling intellectual as ‘alt-right’ to demonize them. Therefore I have addressed both fringes.

  3. David Turnbull says

    “Their results indicate that attacks on an individual’s credibility as a reliable source may be just as effective as attacking the claim that the person is making. ”

    This is stretching the definition of an ad hominem attack. If I say that X is a nasty person so you should not believe them, that is an ad hominem attack. If I provide evidence that X is a consistent liar so you should not believe them, that is not an ad hominem attack.

    In the particular study which you reference and grossly simplify, credibility in science as reflected by past or present misconducts quite rightly affected judgement of claims. Ad hominems, such as the education of the researcher, did not.

  4. Jason Monez says

    While I do agree that ad hominems are destructive to productive discourse, as David Turnbull pointed out, the examples you provided from that study are not at all examples of actual ad hominems! Research is done by a researcher, and if the researcher’s scientific credibility is suspect then the research is too. An ad hominem of the researcher would be if someone said that because he supported corrupt Politician X then his research is obviously flawed. But it’s perfectly appropriate to question the person’s research if it’s pointed out that there are conflicts of interest or questionable research practices in his past! (I agree that one shouldn’t immediately reject the research upon hearing such claims, but increased skepticism is indeed warranted in such cases.)

    Honestly, I’m really surprised that there could even be a study making such a claim. I intend to take a closer look at it.

  5. Jason Monez says

    To add to what I wrote above, a parallel example would be to point out that it isn’t an ad hominem attack when attacking a person’s character if the person’s character is the very thing that is at issue. For example, if a politician is touting their honesty, it isn’t an hominem to point out their past corruption, since past corruption is directly relevant to the issue at hand (their claims of honesty).

    Similarly, when it comes to someone making claims about their scientific discoveries, it isn’t an ad hominem to legitimately attack the person’s scientific credibility since that has a direct bearing upon the truth of their scientific claims.

  6. I broadly agree with David and Jason that ad hominem attacks – maybe a better term would be ‘ad hominem critique/scepticism – can be justified (if scientists are making bold or broad claims with a shoddy record of scientific integrity – e.g., the recent Wansink claims – their work should be interrogated thoroughly).

    However, I think Elio is correct to call these ad hominem. If you’re judging a claim because of the identity and personal characteristics of who is making it, by definition you are critiquing the man (colloquially speaking) and not the argument.

    I can see both sides of this debate, but I think it’s disingenuous to suggest this article isn’t strong because of it (if anything, it supports its central point).

  7. Replying to Jason and David,

    It seems to me the problem is how you are defining an Ad Hominem in comparison to how I and the researchers are. I believe you are suggesting that if the attack is justified then it is not an Ad Hominem, however that would be redefining the term as well as exceeding the limits of its use. An Ad Hominem is simply an attack on the person instead of the idea they possess. The attack is irrelevant because by discrediting the person, you are not attacking the claim directly. It is therefore a logical fallacy.

    Even if you believe it is justified to question a researcher or politician’s argument because of their motives or character, it is a logical fallacy, albeit you may consider it a justified one. It’s technically not attacking the claim on its own merit.

    Whether or not you believe you should be skeptical of a claim if you distrust the author is fine, but if you apply that distrust to attacking the single argument, then it is a fallacy. I believe you are attempting to suggest it is not a fallacy if that person is inherently untrustworthy, however that actually brings us to the point of the article. A dishonest person can still speak the truth and by condemning someone because someone else has suggested they might be dishonest can lead to a very slippery slope.

    At worst, we have a disagreement on if it is still logical if you attack the person and feel its justified, but I would not say that either I or the authors of the paper have grossly misrepresented the definition. I would welcome a counter article if you feel this strongly about it.

    • David Turnbull says

      My claim is that you are using a truncated version of the definition of ad hominem: attacking the person making the argument, rather than the argument itself.

      The full definition is ‘attacking the person making the argument, rather than the argument itself, when the attack on the person is irrelevant to the argument the person is making.’

      See, for instance
      for a layman’s discussion or
      for a more rigorous discussion.

    • Jason Monez says

      Elio, I see your point, but I think the point still stands. Yes, technically one can still call it an ad hominem, but if it is justified then how can it be fallacy? A fallacy is by definition something misleading, deceptive, erroneous etc. If it’s a justified line of reasoning, it isn’t a logical fallacy.

      You’re correct that you didn’t mischaracterize the technical definition of ad hominem. I’d suggest that it would be more accurate to say you mischaracterized the context of when an ad hominem critique is a logical fallacy that should be avoided.

      • akvadrako says

        Hi Jason,

        “logical fallacy” is a category of arguments where it’s not immediately obvious they are making a rational argument. Most of them are effective because we have evolved to treat them as effective, so there very well could be a hidden rational basis.

        It’s kind of an economic term, where “rational” means trying to maximise your returns within the confines of the experiment. People don’t act that way because there is a larger context to understand.

        So being a “logical fallacy” doesn’t mean it’s wrong.

        • Jason Monez says

          I beg to differ. My understanding is that it’s exactly the opposite of what you’re describing. Logical fallacies are arguments where it’s not immediately obvious they are making an irrational argument, and thus easy to fall for.

    • Sapere Aude! says

      If I may butt-in, I think Elio’s post doesn’t quite get to the heart of the objections raised by Jason and David.

      There are occasions where an ad hominem is not necessarily a logical fallacy. For an argument to hold, it must both be logically valid and its premises must be sound. Validity alone is not enough. E.g. the argument

      Premise 1: I like pens
      Premise 2: Pens are animals
      Conclusion: I like animals

      Is logically valid, but unsound as the second premise is false.

      When you criticise someone’s/some organisation’s credibility, ie by calling CNN “fake news”, you can be implicitly saying that whatever their arguments, there is reason to doubt that the premises of whatever arguments they are making are true (and therefore their argument is unsound). Whether that is the case or not is a separate matter, but I believe it shows that attacking the credibility of whoever made an argument is not necessarily fallacious – the logical structure of the argument is not necessarily being targeted.

    • Iag Teyh says

      Ermm..I think what needs to be noted is that this study cannot write to the generalised theories governing ad hominems.. the data extracted from the participants can perhaps strongly suggest first order effects of reading ad hominems in scientific literature, but as pointed out in the comments, the subject of ad hominems is bigger than this

  8. “This study carries serious implications for discourse beyond electoral politics, and into the sphere of academe and wider public discussion.”

    That’s a red flag. Whether a singly study has “serious implications” beyond the the study setting can only been shown by actually studying these other settings. Otherwise such a claim is just empty rhetoric and betrays a lack of understanding for how science works.

  9. augustine says

    Reading this article it occurred to me that there is a corresponding phenomenon to the ad hominem attack. I don’t know if there is a Latin term for it, but don’t we often trust the ideas or utterances of people who have an honorable reputation (personal and/or professional), thereby *refraining* from attacks on his ideas? What effect does such bias have on discourse?

  10. “Such attacks are now commonplace on both sides of the political spectrum, with journalists and academics categorizing intellectual opponents as Nazis or Communist as a way of stigmatizing an individual in the mind of a casual reader not personally familiar with their views.”

    If only being labeled a “Communist” came with even 1/8 of the stigma that being labeled a Nazi or Facist did then you might have a point about both sides of the isle. As it stands being labeled a Communist can get you a weekly opinion column in the NYT or tenure at an ivy league. The only place a communist has been discredited is in economics departments (but then, there’s always Paul Krugman, so….). Being seen as being on the super far left comes with no downsides in this country except in the political arena in flyover country but no one cares or pays attention to them anyway. (I live in Flyover Red State country btw so I know this for a fact).

    So, what, I think, you mean to say in your article is “ad hominem” attacks can get you labeled alt-right and shatter your reputation. Full stop. There is absolutely no equavalent on the other side, whatsoever. If one even happens to mention that, maybe, diversity might not be the end-all be-all and that, maybe, it really is OK to be white then you might as well join up with the Third Reich as far as mass opinion means anything. I guess Hitlers laughing his a$@ off in the grave for how much even the hint of his name or legacy can bring down any decent person or idea, truth doesn’t matter at all (Goebbals would be so proud).

  11. I think the problem Dr. Martino elides over is that honest debate has never been possible after the actors have drawn lines and picked sides. That is where we are now.

    The western liberal world order is in crisis and the ad hominems Martino hears are not so much insults or fallacies as they are emerging party labels like “leveller” and “grandee,” or “commie” and “imperialist stooge.” Those on your side are the “well-affected” and those on the other side are the “malignants.” That is all the rank and file need to know.

    Both sides are now most concerned about winning or keeping power and so propaganda aimed at defining the broad outlines of the struggle is much more import that polite debate on narrow issues.

    It is likely such a debate is impossible because the two sides of a highly polarized polity will never agree on any set of axioms or definitions, which is necessary for a debate. Neither will either side concede victory to the other as long both sides firmly believe that they can not afford to lose the debate.

    The only good thing that will emerge is that neo-liberal imperium that began in 1968 will fall.

    • Sapere Aude! says

      Unfortunately I believe you are right. It seems as if the dominant elites have largely lost interest in attempting to defend their position through superior argumentation and are rather now moving (in fact, have been moving for a while) to suppress dissent. But I doubt they will ultimately succeed.

  12. Pingback: Weekend Reading | Common Sense and Whiskey

  13. Pingback: Sweet release | Me Stock Broker

  14. Pingback: Sweet release – Courtier en Bourse

  15. Pingback: Sweet Relief: Eric Crampton on the Sugar Tax – The Science Page

Comments are closed.