'Pregnant woman' beat 'birthing person'. 'Breastfeed' beat 'chestfeeding'. In a major new study, people preferred several words language guides have been telling them to drop.

Researchers from MacEwan University in Canada and Stetson University in the US recruited 616 American adults to test whether these swaps actually help. They compiled 175 pairs of 'harmful' words and their recommended replacements for participants to rate.

'Participants view the harmful language as generally acceptable,' the researchers concluded, so long as the words were not outright slurs or pejoratives. The guides behind the swaps come from institutions including Stanford and the American Psychological Association.

Study Puts Words to Test

Some guides suggest 'primary bedroom' instead of 'master bedroom' over alleged ties to slavery, and 'pathway' instead of 'pipeline' amid pollution concerns. 'Blind review' was flagged as potentially ableist too, with 'anonymous review' offered instead.

'Tone-deaf' was also flagged for using a physical disability as a metaphor for a lack of awareness. 'Unenlightened' was the guides' suggested substitute for that phrase.

Participants rated each flagged term for familiarity, acceptability and the anxiety it triggered. Most scored as highly acceptable, and several beat their recommended replacements outright in preference ratings.

'Hispanic' beat 'Latinx', and 'addicted' was preferred over 'hooked'. 'Victim' also beat the suggested alternative, 'person who has been impacted by', researchers noted in their published findings.

No Rise in Anxiety Found

A second group of 334 adults then read one of two short stories, one using flagged terms and one using suggested alternatives, before completing surveys. The surveys measured comprehension, emotion and anxiety directly.

Anxiety levels showed no measurable rise between the two groups. Reading the 'harmful' version did not increase anxiety, reduce belonging or dampen positive emotion, the researchers reported.

People from marginalised groups were not more negatively affected by the flagged language than other participants, the study found. Only those scoring high on moral elitism and a need for recognition reacted negatively at all, the researchers added.

Separate research on strong language backs the pattern up. A Pakistani study of 253 adults found regular profanity use correlated with significantly lower stress, anxiety and depression scores, with neither age nor education showing any link.

A 2025 American Psychological Association study likewise found swearing eased inhibition and boosted confidence during physical exertion. Researchers described both effects as closer to a coping mechanism than a harmful habit.

Verbal Abuse in Children

Harvard Medical School research offers a sharper contrast. It found chronic parental verbal abuse, not everyday word choice, was linked to disruption in a brain pathway called the fornix, which is associated with anxiety.

That study compared imaging from 16 adults who experienced verbal abuse against 16 who had not, after screening more than 1,200 people from the community. Disruption severity tracked closely with how severe the abuse had been.

Researchers described this as the first evidence linking ridicule and humiliation to altered brain connectivity. Children's brains, they suggested, may 'turn down the volume' on distressing words as a coping response.

The distinction matters, the offensive-language study authors stressed. Their findings should never excuse racial slurs, ethnic epithets or other genuine pejoratives, they wrote in their conclusions.

'Our findings do not apply to that category of words or phrases,' the authors cautioned, drawing a clear line between everyday terms and true slurs.

Overall, the results suggest flagged terminology like pipeline, breastfeeding and blind review carries little anxiety-inducing weight. Real psychological harm, the evidence suggests, lies elsewhere entirely.

The study authors argued there is little reason to treat such word choices as intentionally harmful. Punishing people socially for using them, they concluded, is largely unwarranted.