I was reading the UK Autocar this week-end, and came across this:

“My ghast is well and truly flabbered.”

This stopped me in my tracks on two counts: (i) I could never say it, let alone write it; and (ii) where did that h in ghast come from? I couldn’t say it because what I would say is My flabber is/was well and truly gasted. Obviously. But what if my native-speaker intuition was at odds with actual usage?

A search for flabbergasted in the OED Online takes me to flabbergast, as verb and noun, with the only spelling options being flaba- and flaber-, along with the derivative flabbergastation. There is no separate entry for flabber, but there are noun, verb and adjective entries for gast, largely with the meaning of “fright” or “fear”. The only entry for ghast was as an archaic or poetic form of ghastly.

A search at corpus.byu.edu on its GloWbE (Global Web-Based English) corpus of 1.9 billion words (from 2012–2013) yields my, ghasted and gasted, as the top three collocates of flabber, and flabber, my and was as the top three collocates of gasted; flabber and my (and the exclamation mark) were the top three collocates for ghasted. However, even the most frequent collocate occurred only three times in this huge corpus.

So, I thought I would do a Google search on the four possible strings based on my flabber was gasted, which yielded the following:

my flabber was gasted       =  325

my flabber was ghasted     =  109

                                    T          =  434

my gast was flabbered       =  184

my ghast was flabbered     =  850

                                    T          = 1034

(I’m assuming that the duplication of results that you get in a Google search would apply to all four listings.)

So we have a 2.4:1 ratio in favour of the nominal being gast/ghast, and a 1.9:1 ratio in favour of the spelling ghast.

So, do we simply have here more variants than I am familiar with, or are they, all, simply, errors? And where does this leave the standing of my native-speaker intuitions?