LINGUIST List 4.780

Thu 30 Sep 1993

Disc: Reanalysis

Editor for this issue: <>


  1. Larry Horn, Re reanalysis: the spittin' image

Message 1: Re reanalysis: the spittin' image

Date: Tue, 28 Sep 93 10:21:02 EDRe reanalysis: the spittin' image
Subject: Re reanalysis: the spittin' image

The recent discussion of metathesis, metanalysis, and reanalysis evokes our
extended colloquy, almost exactly a year ago, on that subset of reanalysis
due to folk etymology--the category of "pullet surprises". One of
the all-time pullet surprise winners, along with the doggy dog world and the
devil-make-hair attitude, is 'spittin' image'. The standard story, as Mike
Kac mentioned during last year's exchange (citing William Safire), is that
the earlier 'spit and image' had become opaque with the loss of the relevant
meaning of the nominal 'spit', and speakers reanalyzed the expression as if it
contained the participle, hence 'spittin(g) image', which is now frequently
seen in print. (The meaning that might be associated with expectorating
likenesses isn't all that transparent either, but let's leave that aside.)
 Now it's clear that the source of the nominal is the trope involving the
verb, as indicated by the following OED citations:
 (1690) We are of our father the devil, like him as if spit out of his
 (1788: Grose, Dictionary of the Vulgar Tongue) He is as like his father
 as if he was spit out of his mouth; said of a child much
 resembling his father.
Slightly later, the nominal emerges, generally in the phrase 'the very spit
of'. The OED provides citations for this meaning of 'spit' (='the exact
image, likeness, or counterpart of') from 1825 ('a daughter...the very spit of
the old captain') and 1836 ('You are a queer fellow--the very spit of your
father'), and somewhat later it appears that 'spit' ceased being, as it were,
a very-polarity item, occurring as well in the collocation 'spit and...':
 (1859) the very spit and fetch of Queen Cleopatra
 (1895) She's like the poor lady that's dead and gone, the spit an' image
 she is.
 Now, of course, it's pretty much ONLY the 1895 use that survives: As
Webster 3 notes, 'spit' in the meaning 'perfect likeness' is 'usu. used in the
phrase "spit and image" <the spit and image of his father>'. But, as just
electronically suggested to me by Debra Halperin Biasca, are we sure this is
really 'the spit and image'? How about 'the spitten image', where 'spitten'
is an instance of the (dialectally attested) past participle of 'spit'? (As
in the anti-cloning ordinance: Thou shalt make no spitten images.) One
argument for such a derivation is the parallel use of the participle in French:
my dictionary cites 'C'est son portrait tout crachE, c'est lui tout crachE' as
'fam.' for 'c'est son portrait tres rassemblant', i.e. his spitten image.
Note also that even the 1895 citation above is homophonous, as transcribed,
with the dialectal past participle as opposed to the conjoined nominal.
(Crucially, it's spit AN' image, not spit AND image.)
Now if Biasca's conjecture is correct, that would still leave 'spittin'
image', with the present participle, as a folk etymology, but with a different
source; this time the opacity arises because a given speaker is unfamiliar
with the dialectal PAST participle 'spitten'.
 One final wrinkle: Caroline Heycock tells me she was informed by someone
a while back that the TRUE source of 'spit and image' is as a corruption of
'spirit and image': a nice instance of second-order folk etymology,
right up there with the Welsh rarebit and the journey cake.
 Can anyone shed any light on the viability of the spitten image?

 --Larry Horn
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue