8

BACKGROUND

Cambridge Grammar of the English Language has this portion regarding backshift on page 154:

  • Backsift not confined to indirect reported speech

We introduced a concept of a backshifted preterite with an example involving indirect reported speech, [10]. It is important to emphasise, however, that the backshifted preterite is used much more generally than in reported speech. Compare:

[21] i This meant that Jill had too many commitments.

ii That Jill had too many commitments was undeniable.

(Bold emphasis mine.)

The that-clause of the former example sentence functions as a complement to the verb "meant", whereas that of the latter the subject of the sentence. These are clearly outside the scope of reported speech. So CGEL does say the backshifted preterite (= past tense) can go beyond reported speech but does not clearly say how far it can go. In fact, CGEL fails to specify whether the backshifted preterite is applicable to other types of subordinate clauses than those mentioned above.

EXAMPLES FOR QUESTION

So I have laid out three example sentences as follows. (Boldfaced are the preterites that CGEL might--but fails to specifically--recognize as "backshift".)

(1) I assumed you were the type who kept your promises.

This one was borrowed from an earlier question. The clause "you were the type" is a complement to the verb "meant", where "were" is recognized by CGEL as "backshift"; the clause "who kept your promises" is a relative clause, where "kept" is not recognized by CGEL as "backshift".

(2) She was surprised that he still loved her.

The clause "he still loved her" is a complement to the adjective "surprised", where "loved" is not recognized by CGEL as "backshift".

The way people experience their iPhone has always started with the display.

(3) So we wouldn't introduce a larger display until we could make one that was great.

This is part of what an Apple's senior VP has said in a promotional video of iPhone 6. The clause "we could make one" is the complement to "until", and the clause "that was great" is a relative clause, where neither "could" nor "was" are recognized by CGEL as "backshift".

QUESTION

And the big question is whether any of the boldfaced verbs can possibly be subsumed under "the backshifted preterite" as defined in CGEL?

EDIT

CGEL at page 153 presents "conditions for backshift" in [13], along with relating examples in [14] and [15].

A backshifted preterite can occur when either of the following conditions obtains:

[13] i. The tense of the matrix clause is past.

ii. The time of the matrix clause situation is past.

The question of what types of subordinate clauses are covered in [13] hinges on the meaning of the term "backshifted preterite" at this point in CGEL (at page 153). Indeed, CGEL has this term defined at the bottom of page 151 leading to the top of page 152, as follows:

...the term backshifted preterite is intended to suggest this change from an original present to a preterite. We retain the traditional terminology for its mnemonic value, but emphasise that backshift is not conceived of here as a syntactic process: we are not proposing that had is syntactically derived by changing the present tense of an underlying have into a preterite. The issue to be considered is simply what the preterite means in this construction.

(Only the latter bold emphasis mine.)

The above-mentioned "this construction" at page 152 refers to indirect reported speech, and there is no change in the meaning of the term "backshifted preterite" between [10] of page 151 and [13] of page 153, in terms of in what construction it is being discussed. Therefore, it is clear that the meaning of the term "backshifted preterite" in [13] is limited to indirect reported speech, and that the types of subordinate clauses covered in [13] are at best limited to content clauses functioning as a complement to a verb of the matrix clause.

In fact, this is corroborated by the first boldfaced sentence (on page 154), which is shown at the top of this question. If the conditions set forth in [13] of page 153 were to cover other types of subordinate clauses than content clauses functioning as a complement to a verb of the matrix clause, CGEL would not have inserted, only after [13], the section of "backshift not confined to indirect reported speech". And even then, this section, as presented at the top of this answer, doesn't really go far beyond content clauses functioning as a complement to a verb of the matrix clause. The only expansion shown in this section is that the content clause may function as the subject of a verb of the matrix clause, as in [21ii] (That Jill had too many commitments was undeniable.).

Community
  • 1
JK2
  • 6,553
  • It seems to me that you might have made some surprising claims of what CGEL says and doesn't say. Could you please insert some supporting evidence for your claims? – F.E. Oct 30 '14 at 19:58
  • It seems that you are interested in understanding the time stuff w.r.t. backshift. Perhaps if you consider a backshift preterite as being similar to a modal preterite, w.r.t. the time related stuff, that that might make things clearer for you. Just a thought. :) – F.E. Oct 30 '14 at 20:01
  • The stuff in CGEL on page 153, related to [13-15], seems to be directly related to your question. – F.E. Oct 30 '14 at 20:05
  • The answer to the question posed in your title is "yes". – F.E. Oct 30 '14 at 20:08
  • @F.E. Please see the "edit" for the "supporting evidence" you've requested. After reading the edit, if you still believe that the answer is "yes", could you please provide me with some explanations in a form of answer so I could award you the bounty and get this over with? :) – JK2 Oct 31 '14 at 03:08
  • the clause "who kept your promises" is a relative clause, where "kept" is not recognized by CGEL* as "backshift".* <== I disagree. Refer to page 153, [13]; and also refer to page 156, [25]. – F.E. Oct 31 '14 at 05:23
  • where "loved" is not recognized by CGEL* as "backshift".* <== I disagree. Refer to page 153, [13]. – F.E. Oct 31 '14 at 05:25
  • where neither "could" nor "was" are recognized by CGEL* as "backshift".* <== I disagree. Refer to page 153, [13]. – F.E. Oct 31 '14 at 05:28
  • The above-mentioned "this construction" at page 152 refers to indirect reported speech, <== I disagree. From the context, I think the authors meant a backshift construction, that is, a construction that involves a backshift preterite. – F.E. Oct 31 '14 at 05:30
  • Therefore, it is clear that the meaning of the term "backshifted preterite" in [13] is limited to indirect reported speech,** <== I disagree. The authors provided examples that weren't "indirect reported speech". For instance, page 152, [11]. – F.E. Oct 31 '14 at 05:33
  • The authors even wrote on page 154: It is important to emphasise, however, that the backshifted preterite is used much more generally than in reported speech. <== That basically says it in a nutshell. – F.E. Oct 31 '14 at 05:37
  • I probably don't have the current opportunity to write an ELU answer-post which would analyze one of your examples by using the time stuff (T_o, T_r, etc.). But you could try to do that yourself by using the information on page 154, [18-20]. – F.E. Oct 31 '14 at 05:41
  • @F.E., I find it surprising for you to claim that examples (1) through (3) are somehow covered in CGEL, when they clearly are not. At any rate, the condition set forth in [13] of page 153 in and of itself does NOT cover "any" types of subordinate clauses. Because if it did it should also cover the second conditionals, which CGEL categorizes as "modal preterite" as opposed to "backshifted preterite". – JK2 Nov 04 '14 at 06:18
  • "… when they clearly are not." <== I give up. – F.E. Nov 04 '14 at 08:27
  • For example, you say "loved" is covered by CGEL. However, nowhere in CGEL is an example of backshift preterite in a relative clause, as in "loved". – JK2 Nov 04 '14 at 08:37

3 Answers3

2

Could all the boldfaced verbs of these examples be subsumed under
"the backshifted preterite" as defined in CGEL?

If you want to subsume them, feel free to do so. They are all grammatical and colloquial English.
In complements -- and all of these clauses are complement clauses -- some subject complements, and others object complements -- except the last, which starts with an unbolfaced preterite-form modal (wouldn't) and continues with another (could) and a preterite in a relative clause (that was great).

Though it's not necessary to label tenses as "backshifted" or not. These are preterite, certainly.
And if one has a usable definition of "backshifted" -- i.e, a definition that can be applied to any case and will always distinguish "backshifted" forms from non-"backshifted" forms -- no problem.

However, it appears that you don't in fact have such a definition (or perhaps CGEL doesn't),
because this question is about what it means. If there's a definition, apply it.
If there isn't, the term is useless. Take your pick.

John Lawler in exile
  • 161
  • I'm not really following what you're saying in that these are all "normal" preterites. If they were, they should be referring to the past time compared to the time of utterance. But are they really? – JK2 Oct 28 '14 at 11:11
  • Not at all. First, most of these are generic. Second, tense is not time; the Present rarely refers to the present time, and the Past can refer to practically any time, depending on what it's controlled by. In these cases, most of them are complements, so the details of the tense are determined by the matrix predicates. – John Lawler in exile Oct 29 '14 at 06:53
  • When you said "the details of the tense are determined by the matrix predicates", do you mean that the past tense of the boldfaced verbs was due to the past tense of the verbs in the matrix clauses? – JK2 Oct 29 '14 at 07:05
  • @JohnLawlerinexile +1 So-called backshift in reported speech is a bogus phenomenon and a bogus concept. It's also something, that, for example, no language students ever have make mistakes with (i.e. using the correct tenses in relation to reported information) - until some numbskull language teacher tries to teach them about it. As soon as this happens, all kinds of problems, confusions and errors start to proliferate. This in itself is evidence that it's all junk. Frankly, the whole concept of reported speech is entirely wonky too. I've never met anyone able to use this term consistently. – Araucaria - Him Oct 31 '14 at 13:51
  • @Araucaria: Which is bogus, backshift or reported speech? Are you saying that the backshifted preterite set forth in CEGL is bogus? Or that reported speech in the traditional grammar is? – JK2 Nov 05 '14 at 02:29
  • @JK2 Backshift in conditionals for example, as set out in CGEL, is a real phenomenon. However, I don't believe that backshift in reported speech is. It's just the standard use of the praeterite to describe past situations and events, imo. The special justification given by H&P regarding must and ought to would work if these were confined to reported speech - but as H&P say, they aren't! Which kind of undermines the justification in the first place ... That's my take anyway ... :) – Araucaria - Him Nov 05 '14 at 02:35
  • @JK2 ... but it would be pretty difficult to come up with a coherent definition of reported speech anyhow. No one I've ever read or spoken to has ever been able to use the term consistently, unless they made the definition so narrow that it was useless ... Your post inspired this question here - but haven't managed to get many answers yet:) Reported speech. Actually that was my second attempt... – Araucaria - Him Nov 05 '14 at 02:44
  • @JK2 ... Here was my first Reported speech1. – Araucaria - Him Nov 05 '14 at 02:44
  • @Araucaria: What do you mean by "the special justification regarding must and ought to"? – JK2 Nov 05 '14 at 02:46
  • @JK2 p.152 H&P state that there are two reasons why the backshifted preterite 'can't be subsumed under the basic use of the preterite'. The first a) is about 'modal remoteness' as well as reported speech and subordinate clauses. The second, b), is basically about ought to and must being able to be used with past time reference in backshifted reported speech, when they can't freely occur elsewhere with past time reference. – Araucaria - Him Nov 05 '14 at 03:00
  • @Araucaria: I find those two reasons plausible. And I wonder what you mean by "as H&P say, they aren't!" Are you attacking their second reason (b)? – JK2 Nov 05 '14 at 03:13
  • @JK2 Yes, I am. If It was clear we must leave is not a form of reported speech, then what is the special justification for saying we need backshift to accommodate must here? Why don't we just say we that must is restricted in its usage in relation to past time? If reported speech does not exist then what H&P are saying here is that we need backshift to explain must, but actually they're using must to explain backshift, it's circular. Their a) arguments are not good either ... – Araucaria - Him Nov 05 '14 at 03:25
  • @JK2 ... . For example, they argue that preterite forms are used in subordinate clauses even when there is no 'modal remoteness' depicted in those clauses. However, when there actually clearly is no 'modal remoteness', we are often completely blocked from doing so: He wishes he was as tall as I was now is definitely not ok. They also, quite unjustifiably, try to maintain that 'modal' backshift is about counterfactuality or unlikelyhood. There is no evidence to back this up ... – Araucaria - Him Nov 05 '14 at 03:30
  • @Araucaria: You can't say, "We must leave by noon yesterday". But you can say, "We knew we must leave by noon yesterday". So how can you explain without using the concept of 'backshift' that the latter is possible when the former is not. – JK2 Nov 05 '14 at 03:33
  • @JK2 Well, there's a number of possibilities and I don't claim to know for sure:) But given that It was clear yesterday that we must leave it doesn't seem to have much to do with reported speech. Here's one theory though. When speakers use must or should part of the meaning is that the speaker or some other entity is decreeing that this is necessary. So if Bob says you must leave he is saying something like I decree that it is necessary/imperative that you leave. No Bob can't say I decree that it's necessary that you left yesterday any more than he can say ... – Araucaria - Him Nov 05 '14 at 03:52
  • @JK2 ... You must leave yesterday. So there must be some kind of actor decreeing or invoking the necessity at the same time as the necessity. That decreeing entity must be implied and is usually the subject of the matrix clause. But notice that this is not always necessary. It was dark, it was minus fifty degrees and we were going to die. We must leave at once or face the consequences. This is ok, because the decree-er of the necessity is obviously ourselves at this past time. We therefore don't need a matrix clause for the must to be embedded in. Now that's just ... – Araucaria - Him Nov 05 '14 at 03:57
  • @JK2 ... a theory, but it doesn't require the invention of a whole special type of tense usage. Some ideas vaguely related to the mad theory I just expounded were suggested by Joos, whom Tunny talks about under their post below ... I apologise for unleashing my madcap ideas on you here! – Araucaria - Him Nov 05 '14 at 03:58
  • In your example "He wishes he was as tall as I was now", "as I was now" is subordinate to "he was as tall", and not directly to "he wishes". One plausible way to explain that "was" here is not possible is to say that the backshifted preterite "was" is not applicable to this particular type of subordinate clause. – JK2 Nov 05 '14 at 04:13
  • @JK2 Yes, but it's directly embedded in the complement of was. H&P use doubly embedded clauses too in their justification: I wish he realised that she had* too many commitments* Compare that with * I wish he realised that right now she was ill. – Araucaria - Him Nov 05 '14 at 04:25
  • 1
    I'm afraid I don't agree that the latter is ungrammatical. – JK2 Nov 05 '14 at 04:33
  • 1
    In fact, CEGL says that the time of her having too many commitments is the present in its most salient interpretation. So I take it that inserting "right now" wouldn't necessarily render the CEGL's example ungrammatical per se:"I wish he realised that right now she had too many commitments." Don't you agree with me on this? And this "doubly embedded clause" (she had too many commitments) is to be distinguished from such a clause as "as I was now" in your example. – JK2 Nov 05 '14 at 06:28
  • @JK2 Actually, yes I do agree that those last two are grammatical. I shouldn't try and pull my grammaticality judgements out of the hat at four in the morning! However, how about "I wish he realised that tomorrow she was going to France"? Actually this is a bit of a disctraction, It's not the main objection to H&P's argument in a) - which is that it relies on an interpretation of 'modally remote' backshift as indicating counterfactuality, which, imo, it doesn't. – Araucaria - Him Nov 06 '14 at 09:55
1

On my reading of Chap. 3, sec. 6.2, Huddleston & Pullum intend that backshifted preterite is a use of the preterite that occurs almost always in embedded clauses whose matrix clause meets either of the conditions in [13i]. Its meaning is to locate the event in the subordinate clause as simultaneous to the reference time established in the matrix clause.

My hunch is that if we have to classify the OP's examples, (1) and (2) contain backshifted predicate (on H&P's definition), whereas (3) uses predicate to indicate modal remoteness. The question does makes me wonder, though, whether the predicate might fulfill both backshift and modal remoteness function in certain contexts.

1

One problem with the concept of backshifting here is that it is based on Huddleston & Pullum's's belief, shared by many writers on tense, that "The general term tense applies to a system where the basic or characteristic meaning of the terms is to locate the system, or part of it, at some point or period of time" (H & P p.115).

Martin Joos (1960.121), The English Verb, Form and Meanings suggests that a more accurate idea of tense might be to consider the so-called present tense the actual tense and the so-called past tense the remote tense. "The latter name fits the meaning precisely. The modern English remote tense has the categorical meaning that the referent (what is specified by the subject-verb partnership) is absent from that part of the real world where the verb is being spoken."

Joos talks about two types of remoteness, in time and reality. My own work on the subject suggests three types, time, reality and directness.

Let's take your three sentences:

(1) I assumed you were the type who kept your promises.

(2) She was surprised that he still loved her.

The way people experience their iPhone has always started with the display.

(3) So we wouldn't introduce a larger display until we could make one that was great.
[These words, at 1.30 in (this video), suggest a meaning similar to "We were not willing to introduce a larger display until we were able to make one that was great".]

Let's also consider a fourth possibility;

At the moment we are not capable of introducing a larger display. This does not worry us because, even if we were capable of producing a crude large display,

(4) We wouldn't introduce a larger display until we could make one that was great.
[These words, identical to those of (3), apart from the deletion of 'so', now have a future hypothetical rather than a past-time factual meaning.]

There is no need to posit any form of 'backshifted preterite' for any of the underlined remote tense forms. In each case, the speaker/writer has chosen to make *the referent (what is specified by the subject-verb partnership) absent from that part of the real world where the verb is being spoken.

tunny
  • 4,808
  • In each of the examples, could you tell me what the "referent" is? – JK2 Nov 04 '14 at 03:08
  • @JK2 - As Joos explained, the referent is "what is specified by the subject-verb partnership". In the first sentence therefore 'I assumed' refers to the speaker and his/her mental act of assuming. – tunny Nov 04 '14 at 06:26
  • @ mari-Lou - I have now written out the names in full in the first sentence. – tunny Nov 04 '14 at 06:28
  • @tunny, I agree with--and appreciate--your distinguishing (3) and (4). Do you think that "could" and/or "was" in (3)'s context may be in the form of "can" and/or "is"? What about (4)? – JK2 Nov 04 '14 at 06:40
  • We wouldn't introduce a larger display until we could make one that was great. – tunny Nov 04 '14 at 07:30
  • @K2 - Sorry. I added my last comment before I had finished it. "We wouldn't introduce a larger display until we could make one that was great." In (3) it is possible to distance the willingness idea of idea of introducing the larger display: "We wouldn't introduce .... until we can make one that is great". However, in (4) the reality of all the ideas needs to be distanced. – tunny Nov 04 '14 at 07:42
  • And in (3) I think it would also possible to say "We wouldn't introduce...until we could make one that is great" wouldn't it? – JK2 Nov 04 '14 at 08:34
  • Yes, it's possible. – tunny Nov 04 '14 at 09:36
  • Just a thought, but under CGEL's categorization of "ordinary preterite", "backshifted preterite" and "modal preterite", (3) belongs in "backshifted preterite" while (4) in "modal preterite". Although you say there is no need for making the category of "backshifted preterite", without the concept it would be hard to figure out whether the present tense "can" and/or "is" would be possible in (3), wouldn't it? – JK2 Nov 05 '14 at 01:32
  • Would you please answer my question above? – JK2 Nov 06 '14 at 05:47