zero vs. "that" relatives

Arnold Zwicky zwicky at STANFORD.EDU
Fri Dec 26 16:01:57 UTC 2008

On Dec 25, 2008, at 9:33 AM, Wilson Gray wrote:

> ---------------------- Information from the mail header
> -----------------------
> Sender:       American Dialect Society <ADS-L at LISTSERV.UGA.EDU>
> Poster:       Wilson Gray <hwgray at GMAIL.COM>
> Subject:      Re: zero vs. "that" relatives
> -------------------------------------------------------------------------------
> FWIW, I prefer the form with "that." I have the *impression" that
> "that" is used more often in BE than in sE. I'm willing to admit that
> I could be wrong about that. Maybe it's just that *I* prefer the
> "that" forms. But my preference for "that," IMO, is based on my
> underlying BE grammar. And, given that more sE speakers than BE
> speakers exist and are more likely than BE speakers to be posting to
> the Web, IAC, that there should be fewer examples with "that" than
> without "that" is to be expected.

i wasn't claiming that i prefer the zero variant (when it's available)
*in general*, only that i prefer it in the particular construction i
posted about.

the facts about "that" vs. zero in relative clauses are very complex.
to get some appreciation of this complexity, check out some papers by
florian jaeger and various collaborators, available at:

(i'll quote from several of these below).

to start with,

"For most speakers of Standard American English, only finite,
restrictive, non pied-piped, non-
extraposed, non-subject-extracted RCs [NSRCs, for short] can occur
without optional that."

and then:

"A variety of factors seem to influence the choice between that and no
relativizer in these cases.  These include the length of the NSRC,
properties of the NSRC subject (such as pronominality, person, and
number), and the presence of disfluencies nearby."

"... lexical choices in an NP containing an NSRC can [also] influence
whether a relativizer is used.  ... particular choices of determiner,
noun, or prenominal adjective may correlate with exceptionally high or
exceptionally low rates of relativizers."

there's more, but this should be enough to show that introspecting
about your *general* preferences for "that" or zero is just hopeless.
someone's impressions about their general practices are not any kind
of evidence about their actual practices (and, even more strongly,
someone's impressions about the practices of an entire group of
speakers are not any kind of evidence about this group's actual

everyone's inclination is to think about what they'd do in a few cases
and then generalize from that.  thinking about specific examples can
be a useful exercise, but the generalization is utterly worthless
unless it's tested -- in this case, tested by examining people's
actual practices (and that's a non-trivial piece of research).  it
doesn't really make any difference what you *think* you (or other
people) do.

it seems likely to me that individual speakers/writers might have
different overall preferences for "that" vs. zero (all other factors
being held constant), and that groups might also differ in this way.
i don't know of any research on the question, though.  i don't even
know what i do myself.


The American Dialect Society -

More information about the Ads-l mailing list