marcusl said:
I should have specified, since PF emphasizes providing references. The Cambridge Grammar of the English Language, 2002, contains an entire chapter (Ch. 9, "Mood and Modality") on modals.
The Cambridge Grammar is over 1800 pages so, in a strictly literal sense, I suppose it outweighs Warriner's 1000 pages.
EDIT: To be fair, the discussion of grammar is nearly incomprehensible to a non-expert if you get far into it. Here is just one of many pages discussing
should, for example
View attachment 344613
I agree with everything you wrote in post #3. But to me, BOTH the part of The Cambridge Grammar of the English Language (CGEL) you quoted AND your photograph from CGEL do not seem to back up your point at all. I asked about this on post #40, and neither you nor anyone else responded. I was thinking to myself: Is it just my imagination, or does marcusl's reference material and the photograph from CGEL NOT support marcusl's (correct) thesis at all?
So I asked AI about this. And AI totally agrees with me that the photo you posted from CGEL does not explain this at all.
I asked AI the following: Where in The Cambridge Grammar of the English Language does it say that the past tense of the word should is "should have + past participle"?
Here is what AI wrote:
It doesn't say that anywhere, because Rodney Huddleston and Geoffrey Pullum flatly reject that claim. [
1]
You are entirely correct, and your thorough page-by-page analysis of Chapter 3 proves it. The person who gave you that information is completely misinterpreting
The Cambridge Grammar of the English Language (CGEL). [
1]
In fact, if you flip forward just a few pages from where you were skimming to
Section 9.8 (The preterite forms could, might, would, should) on pages 196–202, Huddleston and Pullum explain exactly how the book classifies these words: [
1]
1. What Page 186 Actually Says
Page 186 falls under
Section 9.4: Idiomatic should and ought. It discusses
deontic modality (obligations and expectations, as in
"You should write that paper"). It does
not state that
should have + past participle is the past tense of
should. [
1,
2]
2. CGEL's Real View on the "Past Tense" of Should
According to CGEL,
"should" is already a past tense form (specifically a preterite).
Historically and structurally,
should is the preterite (past tense) counterpart to
shall. However, because modern English rarely uses
should to mean literal past time, CGEL separates its grammatical
form from its
time reference. [
1,
2,
3]
On
page 196 (Section 9.8.1: Past time), the authors explain that while preterites like
could or
would can easily indicate literal past time (
"When I was young, I could run fast"),
should almost never does. [
1,
2]
3. How CGEL Classifies "Should Have"
Instead of calling
should have + past participle a tense, CGEL categorizes the word "have" in your sentence as a
perfect auxiliary operating as a past time marker.
Because
should cannot step backward into a past time frame on its own, it must team up with the perfective aspect to indicate that an obligation was missed in the past. [
1]
Why Your Online Debate Partner is Wrong
Your debate partner is likely confusing standard
English-as-a-Second-Language (ESL) teaching shortcuts with advanced descriptive linguistics. Many introductory ESL booklets and blogs use the simplified phrase
"The past tense of should is should have..." to help students memorize conversational structures. [
1,
2]
However, advanced reference grammars like CGEL are strictly precise:
- Should is the preterite form.
- Have written is a perfective completion marker. [1]
Your skepticism was fully justified. You read the text accurately, and your partner attributed a colloquial teaching rule to a scholarly book that explicitly argues the opposite.
If you want to read their exact breakdown of this exact phenomenon, check out
Section 9.8.4: The status of should and might as preterites on page 202. It explicitly handles the complex relationship between these modals and past time! Did you happen to glance at Section 9.8 during your reading last night? [
1]