Can relative pronouns be omitted in some regions?


“I guess it was Cal asked Lee.” (Aron, born in California)
. . . . . .
“That’s a smell could raise me out of a concrete grave.” (Adam, born in Connecticut)
(John Steinbeck, East of Eden)

In the ‘it is’ or ‘there is’ constructions, relative pronouns can be omitted, says my Korean grammar book. In this context, is the second sentence an example of regional dialect?


Posted 2013-06-14T23:20:26.230

Reputation: 25 811

5@ Listenever: You really do make things difficult for yourself by attempting to understand modern standard usage with texts like this. Steinbeck's narrative style itself is fine (there's nothing particularly unusual about his own words, apart from being a bit dated). But when it comes to reported speech, you should bear in mind that he's not writing books intended to be read by people who speak like his characters. He might quite possibly invent non-standard usages for them, because he writes fiction, not accurate history books. – FumbleFingers Reinstate Monica – 2013-06-14T23:42:28.330

1...also note that more often than not, such "non-grammatical" usages don't reflect "regional dialectal usages" as such. They're characteristic of *uneducated Anglophones with poor linguistic skills everywhere* (and often "casual speech" from people who know perfectly well they're being ungrammatical). – FumbleFingers Reinstate Monica – 2013-06-15T00:17:37.670

4@FumbleFingers I'm going to dissent. 1) Steinbeck's got a pretty good ear: his inventions are grounded in ordinary US speech. 2) In that speech the subject null-relativizer is almost as common as the uncontroversial object null-relativizer. It is (just for instance) frequent in my speech, and I regard it not as ungrammatical but as colloquial. McCawley, Ch. 13, fn 4, notes that it was common in Old and Middle English. As far as I know the use is general American rather than regional. – StoneyB on hiatus – 2013-06-15T00:47:58.403

@StoneyB, The book seems to say the structure as 'pseudo-relatives' (p449-51) "..."less tightly connected" with the preceding NP than restrictive relatives are with their heads.. – Listenever – 2013-06-15T02:53:16.207

1Your sentences qualify as those 'pseudo-relatives'; but my dialect, at least, permits sentences which do not: The guy Ø sold me this said it had been in his family for 80 years. Anybody Ø needs to talk to me knows where to find me. And McCawley isn't willing to go so far as to call such sentences as yours pseudo-relatives. He only 'mentions' somebody else's observation that omission of a subject relative pronoun is common 'in a class of cases that appears to coincide with what I call pseudo-relatives'. – StoneyB on hiatus – 2013-06-15T04:51:34.473

2@StoneyB: It's perfectly common in UK casual and/or uneducated speech too. I wouldn't want to get too bogged down in whether this "subject null-relativizer" is properly described as "ungrammatical", but I really don't think it's a speech pattern anyone should go out of their way to learn. You wouldn't expect an English teacher to include this sort of thing on the curriculum. – FumbleFingers Reinstate Monica – 2013-06-15T13:06:18.457

1@FumbleFingers I'd put it differently: it's not a use they should go out of their way to use, but they do need to learn it if they're going to actually talk to people - or read many interesting books. – StoneyB on hiatus – 2013-06-15T13:13:11.797

@StoneyB: In fact, Steinbeck's is the only instance of "That’s a smell could" in Google Books. I tried looking for the same feature in something a bit more likely to occur, because I'm pretty sure in BrE (dunno about AmE), the position you call a "subject null-relativizer" would often be filled with as (presumably a shortening of such as). I found a few each of he's a man as could/would/should. Not many, but certainly more than I found of the same construction without "as".

– FumbleFingers Reinstate Monica – 2013-06-15T15:59:54.107

@FumbleFingers I think of that as a mostly British usage. OED 1 (As, 24.) suggests that this is a reduced form of such ... as, but I doubt any modern user thinks of it that way. – StoneyB on hiatus – 2013-06-15T16:52:30.230

@StoneyB: Yeah, I use it myself sometimes (when I'm feeling "rustic/plebby"). But I actually think of it as a non-standard substitute for that - I only put the "such as" idea in because I guessed that might have been the origin, but that's certainly not what's in my mind when I use it. – FumbleFingers Reinstate Monica – 2013-06-15T17:31:28.947

@FumbleFingers Pretty much what OED says: “The antecedent such is also replaced by that, those, or entirely omitted, leaving as an ordinary relative pronoun = That, who, which. Cf. Norse use of som. Obs. in standard English but common dial. in England and the United States.” – StoneyB on hiatus – 2013-06-15T17:35:17.897

So apparently as in such contexts isn't a particularly BrE usage then. Interestingly, after ignoring dups there are 3 instances of kind of man as would in GB, and it's easy to see 2 of them are American.

– FumbleFingers Reinstate Monica – 2013-06-15T17:52:52.710

@FumbleFingers One however from 1909, and the other a work of fiction set in the first half of the 19th century. – StoneyB on hiatus – 2013-06-17T11:19:19.073

It might not actually be regional, but to my Midwestern ears, it sounds Southern. – snailplane – 2013-06-20T04:14:26.513



Deleting a relativizer which stands as the subject of a relative clause is not acceptable in formal English, but it is very common in US spoken English. I believe it is to be found in all US dialects, but there are two different processes at work here, and I am not expert enough to say which governs outside my own region:

  • In the Southern dialects I am natively familiar with (and probably in those Afro-American dialects which have 'descended' from Southern speech) this is often a simple deletion, just like the universally accepted omission of a relativizer which stands for an object of a relative clause. It is not, I think, a 'mistake' or an innovation but a dialect survival from Middle and Early Modern English.

  • There is also a phonological elision so extreme that it is practically indistinguishable from frank syntactic deletion. Function words like relativizers are almost always unstressed, and therefore reduced to a 'weak' form. With the weak form of that—which is far and away the most common spoken relativizer—there is reduction in all three components:

    • The initial /ð/ is dropped.
    • The vowel is reduced to /ᵻ/.
    • The /t/ loses its alveolar articulation and is reduced to the glottal stop which usually accompanies word-terminal voiceless stops.

    But if the the previous syllable ends on a vowel or on a consonant which joins with /t/ in an acceptable syllable coda, the reduced vowel is elided altogether—which reduces that to a single phonetic element, the glottal stop [ʔ]. And [ʔ] is not an English phoneme; it isn't actually 'heard' by native speakers.

    In effect, the word that disappears, phonologically.

StoneyB on hiatus

Posted 2013-06-14T23:20:26.230

Reputation: 176 469

+1 particularly for that final section. If I say "You're making a racket [?] could wake the dead", I might on some occasions firmly believe I did articulate something that was supposed to represent the word *that* (though on other occasions I might "deliberately delete" the word). But the extent to which I might "reduce" the sound is so extreme a listener might well have little or no chance of distinguishing my extreme reduction from my outright deletion. – FumbleFingers Reinstate Monica – 2014-10-05T15:55:11.743