Statistical pregnancy

I was surprised last week to come across ‘pregnant’ being used as a direct synonym for ‘significant’; in its metaphorical uses I had always thought of pregnancy as conveying promise or potential rather than established fact or (near) certainty.

The OED has two separate entries for ‘pregnant.’ One traces the origin to classical Latin praegnant-, praegnāns: with child, pregnant, swollen, but surprisingly finds the earliest evidence of its use in English in the figurative sense of ‘Full of meaning, highly significant’, with a back-up quotation from 1402. But as very early evidence of publication bias, I like this one from Godwin in 1601:

Because my proofes are not pregnant ... I will pass him over in silence.

However, there is also a second entry for pregnant, a word ‘Apparently’ from Middle French meaning ‘Of an argument, proof, piece of evidence, etc.: compelling, cogent, convincing; clear, obvious’ – obviously P < 0.05

But the word has apparently (that word again) been associated in sense with the other kind of pregnant from an early date, and the two words are frequently difficult to distinguish, so the OED tags it as ‘Now rare, frequently with punning reference’ to the other kind of pregnant.

As modern usage of significant in the statistical sense is so sadly prone to misunderstandings I think it would be a very good idea from now on to adopt the use of ‘pregnant’ in its place. At least it would make people stop & ask for an explanation of what we mean by that.

And I love the image of Cern swelling with data until, finally, giving birth to compelling, cogent, convincing, clear, obvious proof of a little Higgs boson.

Links
Being Statistically Significant Is Nothing Like Being Pregnant
Exploring the Significance of Pregnancy and Birth
[PDF] Pregnancy Decision Making as a Significant Life Event
At what point did you start looking "significantly pregnant"?
Related posts
Heteroskedacity
Dependency, privilege & bankers bonuses