It is in the news that Laurence Fox is suing Simon Blake and Nicola Thorp for libel after they made allegations of racism.

There has been quite a lot of work in recent years about ingroup bias studied with fMRI. It would seem to me a justifiable conclusion from these studies that most people are a little bit racist if you look hard enough, in that their brains respond to people of the same race differently and more positively than people of different races. Could such an argument, perhaps accompanied by fMRI data of the plaintiff, be a possible defence in such a case?

The sort of work on ingroup bias I am referring to is summarised in this 2018 review:

Functional magnetic resonance imaging studies have revealed different neural responses to perceiving faces of ingroup and outgroup members, and these effects seem to be task dependent. For example, Cunningham et al. (2004) found that when Caucasian participants watched Black and White faces that were presented very briefly (i.e., 30 ms, so that they were barely a flash on the screen), they showed increased activation in the amygdala in response to the Black faces (which was interpreted as an increased emotional response for outgroup faces). However, this effect disappeared when the pictures were presented for a longer time so that they were clearly visible (i.e., 525 ms). In addition, the prefrontal cortex was more active for White faces compared to Black faces in this condition. The authors suggested that the longer stimulus presentation in this latter condition, allowed the participants to regulate the automatic implicit amygdala bias observed in the former condition through increased prefrontal cortex activation, because participants did not want to be biased, or perceived as biased.

Wheeler and Fiske (2005) also found that amygdala responses were influenced by different face perception tasks. Caucasian participants showed more activation in the amygdala during a social categorization task (i.e., a task where participants had to categorize people based on their age) when observing Black compared to White faces. However, no group difference was found in a non-social visual search task (i.e., a task in which people had to detect whether a dot was present on someone’s face), and more amygdala activation for White compared to Black faces was found during an individuation task (i.e., a task in which participants had to think about what the person liked). The authors concluded that: (1) the simple social categorization task resulted in the increased automatic emotional prejudice response for outgroup targets and thus increased amygdala activation; (2) the faces during the non-social visual search task were not processed deeply enough to represent a social target, and therefore no group effect was observed in the amygdala during this task; and (3) the individuation task resulted in deep level controlled processing and thus a suppression of amygdala responses to outgroup targets as seen in the Cunningham et al. (2004) study.

Another brain region modulated by group membership during face perception is the fusiform face area (FFA). Using fMRI, Golby et al. (2001) showed that this region responds more strongly to same-race faces compared to other-race faces. However, in another fMRI experiment it was shown that when White participants were randomly assigned to a novel mixed-race team, they showed more FFA activity for ingroup team (vs. outgroup team) faces regardless of their race (Van Bavel et al., 2008). In a similar follow-up experiment, which also included a mixed-race control condition with faces that did not belong to the ingroup or outgroup, Van Bavel et al. (2011) again showed enhanced FFA activity for ingroup (vs. outgroup) faces regardless of race (but see Ratner et al., 2012). By comparing the ingroup and outgroup conditions with the control condition, they showed that this increase was caused by enhanced FFA activity for the ingroup faces rather than diminished FFA activity for the outgroup faces. Together these fMRI studies on face perception show that people can regulate the increased amygdala response for outgroup faces, and that we activate the FFA more for faces of ingroup members, regardless if they are the same race or not.

3 Answers 3


The defence would be "fair comment" in Canada, or "honest opinion" in the U.K.

A claim that someone is racist would have to be defended under the "fair comment" defence, not the truth defence. There are "many cases in which Canadian courts have determined that 'loose, figurative or hyperbolic' labels, like homophobic, transphobic, bigoted, racist, or sexist are properly characterized as comment, not fact" (Hansman v. Neufeld, 2023 SCC 14, para. 111). Statements that are comments are not less defamatory, but it changes what defence is applicable.

The fair comment defence has the following elements (WIC Radio Ltd. v. Simpson, [2008] 2 S.C.R. 420, 2008 SCC 40):

(a) the comment must be on a matter of public interest;

(b) the comment must be based on fact;

(c) the comment, though it can include inferences of fact, must be recognisable as comment;

(d) the comment must satisfy the following objective test: could any [person] honestly express that opinion on the proved facts?

(e) even though the comment satisfies the objective test the defence can be defeated if the plaintiff proves that the defendant was [subjectively] actuated by express malice.

Elements (a)-(d) must be proved by the defendant on a balance of probabilities. If the defendant is successful that far, the plaintiff can defeat the defence by establishing malice as set out in (e).

Element (b) requires that the "factual basis for the impugned statement must be explicitly or implicitly indicated, at least in general terms, within the publication itself or the facts must be 'so notorious as to be already understood by the audience'" (Hansman v. Neufeld, 2023 SCC 14, para. 99). A person hearing or reading the claim of racism must be given the underlying facts upon which the claim is based so that they can judge for themselves.

In the U.K. the very similar defence of "honest opinion" would be the applicable defence. It is codified at s. 3 of the Defamation Act 2013.

This also requires that the factual underpinning for the comment is made known to listeners or readers, and that the opinion is one that could be honestly held based on those facts. The defence is defeated if the plaintiff shows that the defendant in fact didn't believe what they were saying (in place of the "malice" prong from Canada's fair comment defence).


It depends on the imputation that is being defended. But this particular data is unlikely to help resolve a disputed question of fact.

I have not read the statements that Mr. Fox is upset about. But from that raw matter, and based on submissions by the parties, the judge would be trying to discern the nub of the allegation being made against him. Frequently, the original published material will be a bit wandering and vague rather than containing a brief and precise logical statement. (For example, the defamation case between Rachel Riley and Laura Murray was partly about the interpretation of a tweet by LM that alluded to a quote-retweet by RR, where RR's contribution was two words and two emoji. The poor judge had to figure out whether LM had fairly summarized RR's intent.)

Broadly, when somebody is being accused of racism, the allegation is not "everybody is a bit racist, and Joe Bloggs here is no different". The context is almost always that Joe is being accused of being more racist than other people, in relation to specific acts done or statements made by Joe. It is implausible for someone to come and claim that they didn't mean to single out Joe, but were just trying to make a point about unconscious bias felt by people in general.

So there is a step in the process where, given all the context, the judge will make a determination about what readers of the material would be expected to gather from it. Further, it would have to be defamatory or else there is nothing to defend. To the extent that the allegation relies on a statement of fact, that statement could be defended on the grounds that it is true. That does not apply to statements of opinion, which could be defended as "honest opinion" or in other ways.

Suppose that the allegation against Laurence Fox is of the form "Fox did XYZ thing, which makes him a racist who should be shunned by society". His doing XYZ is a fact (or not) and the interpretation put on it is an opinion.

  • The mentioned research on in-group bias is of little help in resolving a question of fact here. The statements of fact would be about things that Fox said or did. In trying to show that those things really happened, it is not enough to show that they are the kinds of thing which might happen. If he personally is accused of demonstrating in-group bias then it is no use saying that other people's brains lit up in a certain way when inside an MRI machine.

  • If the allegation XYZ is "we put him in a racism-detecting machine and the needle pointed to 'very racist'" then we do have a statement of fact, albeit an unlikely one. The defence could present evidence of the MRI appointment and its results in order to demonstrate that this happened.

  • The "honest opinion" defence requires among other things that it would be reasonable to hold that opinion on the basis of facts known at the time, and mentioned in the original publication (Defamation Act 2013, s.3). In the unlikely event that Fox agreed to go in an MRI machine as part of the legal process, any results would not help the respondent because those results would be new data not available to them at the time of publication. It would also not help him as claimant because the stated opinion might still have been honestly held at the time.

  • For the same reasons of chronology, a post-hoc MRI scan would not be relevant to an allegation like "Because of the way he did XYZ, I think he is so racist that it would show up in a brain scan". The point is whether that statement is the kind of thing that somebody could honestly believe on the basis of XYZ, not whether it is in fact true. A court would also be likely to interpret this sort of statement as simple hyperbole, and find that the underlying imputation is simply that the person is a racist, not that anything specific is going on with their brain.


One would have to point to open displays of racism

What someone thinks inside their own head is not something that another person could have used to draw the conclusion that one is a racist. Or believed one was a daffodil. Or anything else.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .