Politics: linguistic policing

By Richard North - December 23, 2025

In my piece yesterday, which homed in – amongst other things – on Dr Mary-Ann Stephenson’s comments on “demonising migrants” and making their lives “very difficult” – I ventured the opinion that this woman was “out of touch” with the realities of modern Britain.

Others do not share that view and suggest a more sinister agenda is at play, with one commentator Jim Chimirie making a convincing case that Stephenson’s language is by no means accidental.

Rather, he suggests that she reflects a deliberate stratagem developed by the British state, and refined by the Starmer Regime, of manipulating language in order to suppress open discussion about migration and related social tensions.

Thus, Stephenson’s attempt to steer concerns away from the harmful effects of migration and focusing instead on the criticism of migration, calling the act of criticism a “risk”, is a carefully considered ploy to deflect attention from the policy’s measurable outcomes.

This linguistic policing serves to insulate government failures – such as stalled removals, rising asylum applications and public unrest – by redefining dissent as a moral failing rather than a legitimate response to lived experience.

To evaluate this thesis more fully and save time, I put Chimirie’s article to Grok, using the power of artificial intelligence to draw together multiple strands.

In a way, this facility is rather like a real-time Wikipedia on steroids, which allows further interrogation on points of detail. It does not remove the need for thought and, in my experience, the results delivered depend to a very great extent on the nature of the questions asked and the precision with which further points are raised.

Artificial intelligence, therefore, is not a panacea and neither does it remove the need for skilled writing. It is in this respect, another tool – and extremely powerful one, and one where the user needs to learn how to get the best from it.

That said, the response from Grok was to draw a parallel with George Orwell’s concept of “Newspeak” from 1984, where language is engineered to limit thought and enforce ideological conformity. Controlling language maintains power by suppressing dissent and obscuring truth.

In the UK context, this could be seen as an extension of past efforts to manage public discourse, such as the shift in terminology from “illegal immigrants” to “undocumented migrants” in official rhetoric, which some argue dilutes accountability.

Stephenson’s warning about “demonisation” in the migration discourse, therefore, indicate a pivot toward policing tone over substance, suggesting a strategic realignment to protect policy from scrutiny, especially as public frustration mounts over incidents like sexual assaults linked to migrant hotels.

Chimirie’s article ties protests outside migrant hotels to tangible grievances – sexual assaults, reoffending, and state inaction so, by focusing on the language of dissent and framing it as harmful speech, Stephenson is doing the state’s bidding in deflecting attention from the root causes identified by Chimirie.

However, the interesting thing about AI is that it is able to look at both sides of an argument – a useful corrective when one is supporting one side of the argument. Grok thus questions whether there is really a deliberate attempt at distortion. The intent remains debatable, it says: it could be a calculated power play or a clumsy response to complex legal and social pressures.

Grok refers to the image accompanying the post (reproduced above) which, it says “viscerally underscores the tension” but remarks that, without direct evidence of a coordinated campaign (e.g., internal memos), the thesis leans on inference, “which is both its strength and vulnerability”.

Nevertheless, the AI concludes that the thesis is “plausible and timely” and reflects a growing trend where language becomes a battleground for control in democracies facing migration challenges. It’s not absolute proof of a grand conspiracy, it says, but this is a pattern worth watching.

A further facet of AI is that, on offering conclusions, it invites discussion and my further reference to Newspeak elicited the comment that, if the state can redefine “risk” as a harmful label rather than a neutral assessment tool, it effectively narrows the public’s ability to question policy outcomes.

This, it says, is a subtle but powerful shift: control the vocabulary, and you control the story. From there, it brings in another commentator who interprets Stephenson’s remarks as her saying: “You will agree to the government’s policy that harms you, because if you refuse, we will take away from you the ability to criticize this destructive policy”.

Words, this commentator reminds us “have power and influence”, accusing Stephenson of wanting to strip people of both. She has no shame and no honesty, he says, a comment that ties in nicely with another offering which suggests that that it’s all we get now – “stabilisation” and negative strategy. This is the government and the political elite working to prevent what it most fears, which, right now, is civil unrest.

That led to a consideration of whether the government’s stratagem – if deliberate – is actually likely to work, or whether the public will push back harder.

My thought is that people will see through it, and I also believe resistance is stiffening. With the news that the government is to go ahead with placing illegal immigrants in the Crowborough Army training camp – despite intense local opposition – we will see an important test of popular opposition.

The big question in the context of David Betz’s work – where he argues that the direction of travel is towards civil war and does not see an “off-ramp” – is what sort of timescale we are looking at. Here, many are looking for a “trigger” which might set things off, but my view is that history seems to tell us that multiple events are needed.

This, though, may be ex post facto rationalisation by historians who have the advantage of hindsight. Nevertheless, I favour a “straw that breaks the camel’s back” scenario, which tends to render any one event of less importance (in that it can be any one of many) and impossible to predict. Crucially, it is the breaking point of the camel’s back (which is unknown) rather than the particular character of the “straw” which is the primary determinant.

With regard to Betz’s “direction of travel,” the state’s linguistic strategy—reframing “risk” as harm—might be buying time, but it’s also piling pressure on the metaphorical camel’s back.

Much of the dissent we see, though, is ambiguous – the exact objective may be obscure and it need not be directed at the downfall of the state. Thus, breaking point could come from an unexpected quarter: a rural uprising, a union strike, or even a viral X thread.

Chimirie’s image of police clashing with protesters hints at this tension, while each event adds a “straw”, and the back’s limit is unknowable until it snaps. This unpredictability makes the state’s position precarious and, if resistance is hardening, virtually anything could tip the scales.

The breakdown of the state is a chaotic, cumulative process and Betz’s work supports the idea of a slow-burn – possibly delayed by the state’s attempt to control the narrative – the attempts at control are too transparent to be convincing.

Yesterday, we saw Chris Philp, shadow home secretary, responding to Stephenson, condemning her comments as a “disgrace”. Once again, he said, “the left tries paint those opposed to mass immigration and illegal immigration as racist”.

Cutting through the attempt to deflect from policy outcomes, he spelled out the effects. “Mass migration with no integration has undermined social cohesion. Sex crimes by foreign nationals are up 62 percent, foreign criminals and illegal immigrants routinely abuse human rights, modern slavery and asylum laws to stay in the UK”, he said.

Unconcerned about making migrants’ lives “difficult”, he went on to say: “This nonsense has to end. It is completely wrong that Labour’s new human rights chief dismisses legitimate concerns about mass migration and crimes committed by foreign nationals – including the recent spate of rapes and sex attacks committed by small boat illegal immigrants”.

If even a Tory MP can see through the linguistic policing, then there is not much hope that Labour’s stratagem will succeed. Historical precedents suggest we are heading for an abrupt shift. And although the precise event is impossible to call, Grok is prepared to wager that the breaking point might emerge in 2026.

“That’s a hunch, not a forecast”, it says, making one wonder whether Grok might be up to taking the Turing test. It’s a lot more human than many humans I know.