In the Cambridge Grammar of the English Language; Huddleston and Pullum 2002, they make the following qualifying comment:

... reported speech covers the reporting of spoken and written text but also that of unpoken thought. (p. 1023 - bold H&P's)

We can immediately see from this excerpt that reported speech is being used as a technical term to represent a particular linguistic phenomenon, not as a literal interpretation of the two words 'reported' and 'speech'. This is demonstrated by the fact that reported speech is given by these authors to include not only written text, but also unspoken thoughts.

In comments on this thread: https://english.stackexchange.com/questions/205730/what-exactly-is-reported-speech-does-it-really-exist-and-how-do-you-recognise it is proposed by various commentators that to be reported speech, there must first be some speech or thought to be reported. Reported speech, it is claimed, is a report 'of what someone else said' (italics original).

However in their section on indirect reported speech (p.1024), two of the first examples of indirect reported speech given by CaGEL are:

  • Did she say if I'll be invited?


  • Will I be invited, did she say?

Now the answer to both of these example questions (which are the same question framed in two different ways), may well be: "No, she didn't". One thing, for certain, is that the person producing the 'reported speech' here has no knowledge of the original spoken text at all. In fact, they don't even know if there even was such a text in the first place. There may very well have been none.

So, on the basis of the views given by the commentators on the linked-to thread, which do not seem altogether unreasonable (with the caveat that the views are not unreasonable if based on either established practice or authoritative sources), this should not be classed as reported speech. There is no known original speech being reported.

My question is, are the two examples above, examples of reported speech? If so, what are the specific criteria for reported speech which are satisfied by the two examples. I have not been able to find any such criteria in CaGEL. If these are not examples of reported speech, which criteria of reported speech do they fail to meet? - and what authoritative sources can be referred to, to back up this point of view?

Apparently, such problems are easily resolved by recourse to readily available resources, but I have not been very successful. Any help or genuine insights, therefore, would be greatly appreciated!

  • 2
    It seems to me that the speech need not actually have been uttered to qualify as reported speech; it is sufficient to reference speech that is hypothesized as possibly having taken place. This places it in a parallel category to the unspoken thoughts which H&P have already granted the status of reported speech.
    – Erik Kowal
    Commented Nov 3, 2014 at 4:44
  • @ErikKowal +1 Yes, that's an approach to this particular issue. I don't disagree, but a problem there, potentially, is that when we report our own unspoken thoughts, this is just us expressing our thoughts - they have to be someone else's to be Reported Speech, it seems. The other issue is: what is the original hypothesised thought? Is it that they will be invited? or they won't be invited? Bit difficult to pin down, but maybe that's not a problem ... :) Commented Nov 3, 2014 at 10:11

4 Answers 4


The specific criterion for the syntactic construction called reported speech (or indirect speech or indirect reported speech) that is satisfied by the two questions (Did she say if I'll be invited? and Will I be invited, did she say?) is that both contain the reporting verb "say" - either in the matrix clause or in what the CGEL (p1204) calls a "parenthetical, a kind of supplement".

Assuming that John is the asker of the question, he could rephrase it in direct speech as: Did she say: "John will be invited?"

The Oxford Dictionary Of English Grammar, in its entry on reported speech (p361), states: "Reported speech is the same as indirect speech." The ODEG continues: "When we report speech we can use an introductory reporting verb (e.g. say, tell). This is the usual meaning of the term."

In its separate entry on indirect speech the ODEG (p214) states: "The term indirect speech is often used loosely to cover the reporting of thoughts, using an introductory verb of thinking."

The Cambridge Grammar Of English (Carter & McCarthy, p805) extends the scope of indirect speech to include utterances that use a noun phrase:

Speech reports, both direct and indirect, are most commonly made with reporting clauses containing verbs such as ask, say and tell with a reported clause. There are also other, more indirect ways in which people's speech can be reported, by using nouns such as argument, comment, complaint, observation, remark to refer to someone's words.

  • I didn't like his comment that we were spending too much money.
  • Their biggest complaint was that the room was too small.

The following extract from Yule's discussion of the topic in Explaining English Grammar is more relevant as an answer to the OP's original question about what can be regarded as reported speech (which was closed for reasons unclear to me).

Yule (p274) focuses on the semantic differences between direct and indirect speech, noting that:

The effect of backshift in tense (in indirect speech) creates a sense of 'more remote' ... This effect makes the indirect speech forms more like a narrative account of an event ('telling') and distinct from the dramatic presentation of the event marked by the direct speech forms ('showing').

Yule goes on to introduce a third category that he calls "Summarized reports", in which there is a even greater remoteness between what was said and what is reported.

The functional distinction between the dramatic nature of direct speech and the narrative effect of indirect speech is made more extreme when the structure associated with indirect speech is used to summarize a speaking event as a way of reporting it. The difference between what was actually said, as in [8a], and how it was reported, as in [8b], can be quite large.

  • [8] a. "I am waiting here for you. Where are you? You're never on time!"
  • b. He complained about her being late.

The summarized report in [8b] creates an even greater distance between the speaking event and the reporting event. It also results in much greater control being taken by the reporter for the interpretation of the speaking event. There is, then, a conceptual distinction between the three types of reporting formats in English (Direct Speech, Indirect Speech, Summarized Report).

Yule differentiates between the words typically used in the three "quotative frames". For direct speech the quotatative frame includes verbs "which indicate the speaker's manner of expression (e.g. cry, exclaim, gasp), voice quality (e.g. mutter, scream, whisper), and type of emotion (e.g. giggle laugh, sob). It can also include adverbs (e.g. angrily, brightly, cautiously).

The quotative frame in indirect speech tends to include verbs "which indicate the purpose of the utterance (e.g. admit, agree,deny,explain, promise, repsond, suggest). Such verbs present an interpretation by the reporter of the speech act being performed.

The quotative frame in summarized reports includes verbs such as "chat, describe, gossip, speak, talk".

  • +1 Clear and useful. Thanks for including lots of (non-internet properly published) sources! Commented Nov 8, 2014 at 11:32

Quirk et al (A Comprehensive Grammar of the English Language, 1985.1025) prefer the term Indirect speech, about which they say simply "Typically, indirect speech is used to report statements" and (p. 1029) "Our examples have so far been of indirect statements; but all the main sentence types (questions, exclamations, directives as well as statements) may be converted into indirect speech." Nearly all of their 70 or so examples are of words actually spoken, though one, She thought she had told me that breakfast is served between seven and ten reports a thought and words that were possibly not uttered.

Palmer (The English Verb, 1974.43-46) makes no attempt to define the term reported speech, but his examples include These arguments showed that the world is/was round.

Leech (Meaning and the English Verb, 20-4.107) does not clearly define the term, but his examples include You should have asked the mechanic what was wrong and You must have realised they were bluffing.

I agree with Araucaria that in H & P's "... reported speech covers the reporting of spoken and written text but also that of unspoken thought. (p. 1023 - bold H&P's) (p. 1023 - bold H&P's) the term "is being used as a technical term to represent a particular linguistic phenomenon, not as a literal interpretation of the two words 'reported' and 'speech'".

It seems to me that indirect/reported speech (I prefer indirect) can only be a technical term, if for no other reason than that to use it an any other way poses a large number of problems. If we take the 'speech' part literally to refer to words actually spoken, then none of the following sentences is an example of indirect speech:

Did she say that she is/was coming? - The speaker does not know if the words "I am coming" were actually spoken, so cannot be reporting speech.

The chairman will reveal tomorrow that profits this year are 34% lower than last year's profits. - The words have not been spoken yet, so this cannot be a report of words actually spoken.

I thought he was going to sack me. - This is a report of a thought, not of speech.

If he had told me that he was going to resign, I would have tried to dissuade him. - He did not tell me, so clearly I am not reporting words actually spoken.

She texted me that she is/was coming next week. - Written words are clearly not speech.

They may not be reported speech in a narrow interpretation of that term, but they function syntactically in exactly the same way as 'true' reported speech, for example She said that she was coming. The 'rules' about (non-) backshifting, changing first to third person pronouns, changing time expressions (for example today to that day. etc,) are the same. It seems therefore helpful to have one term for all these sentences. Indirect seems to me to be more appropriate than reported, as some of the possible reported/indirect statements are clearly not reported. 'Speech' is clearly not entirely appropriate, but it has been used for a long time. It seems to me that it is acceptable as part of a technical term, so long as teachers and writers using it make it clear that in the technical term the word speech covers thoughts the written word. Whether or not the words have actually been produced need be addressed only if this issue is raised, in my opinion.

  • +1 for addressing the question!! (and the quality of your answer, of course!) Do you really feel that there is ever any 'adjustment' of pronouns as it were. It makes it sound as if there's a conversion process going on: you put the notional direct speech in here and the reported speech some out there. What I mean is if someone says "You stole my peanuts!" and you say "No, I didn't", we don't assume that there's any 'conversion' going on there.It's just the normal use of deixis. Is there a different process going on in reported speech from the normal deictic use of pronouns, do you think? Commented Nov 6, 2014 at 11:50
  • 1
    @Araucaria - I agree that the pronoun usage in both direct and indirect speech is the normal use of deixis, in normal situations, for native speakers and many learners. However, my personal experience suggests that learners (including young native speakers) who are required to transform direct into indirect speech concentrate so much on backshifting that they forget about the pronouns and time expressions. That is one reason I mentioned them. The other is that mentioning them makes a stronger case for regarding all the utterances I listed as indirect speech.
    – tunny
    Commented Nov 6, 2014 at 12:22
  • I was once given some advice by a DELTA tutor never to teach reported speech, because until you do they will never make any mistakes outside of the normal tense mistakes that they make anyway. I have experimented with this and it seems to be true, that if you don't ever teach them reported speech they never stuff it up in their writing or speaking. However, if you start giving them transformations they will. My theory about why this is so, is that we never transform people's speech in this way. We just report the information that they gave us. We barely ever can remember the exact ... Commented Nov 6, 2014 at 12:34
  • ... wording, and rarely do we care. If we did we'd use direct speech (yes, should have used direct and indirect). Often what is considered to be reported speech has no syntactic or lexical connection with what was actually said. So, in response to what was that Michael Moore film about we might say Oh, he just says that we went to war in Iraq over oil. Of course that's just a summary of the information, not the speech ... That insight from students only stuffing things up after you start teaching them reported speech is why I don't really believe in any conversion ideas :( Commented Nov 6, 2014 at 12:39
  • But I definitely do agree that reported speech must have some kind of syntactic properties, if it is not just a trivial term :) Your young native speaker observation is interesting ... Commented Nov 6, 2014 at 12:48

I am hoping you get an answer from a real linguist, because I am unable to find a truly authoritative source to answer your question. I will attempt to provide some insight from my perspective as a native speaker, backed by some less-than-stellar sources. Those sources do cite others, which might be of more use to you.

When conveying what someone (you or someone else) said, you can either quote them verbatim (direct speech) or summarize their message (indirect speech, aka reported speech). (source)

Direct speech:

John asked, "Do you want to go to the mall?"

I said, "It's hot outside today."

Indirect speech:

He asked if we wanted to go to the mall.

I remarked that it was hot outside.

Indirect speech is important as a linguistic concept to explain why there are certain pronoun or verb shifts when you quote someone indirectly.

At least one source (1) argues that indirect speech can also include "future or hypothetical discourse". "Did she say if I would be invited?" or "If I ask her to go to the dance, will she say yes?" But there does not appear to be universal acceptance on that score.

These, at least, still have the same speech patterns as real speech, so intuitively it feels right to include them.

And from there it gets murkier. Once we start including hypothetical speech, should we also include things that are not technically speech at all?

John thought she looked sad today.

He ordered them to bed.

Is this indirect speech? There's no actual speaking involved, but one could imagine as if there were. "Wow, she looks sad today," John thought to himself. "Get to bed this instant!" Sally hollered, pointing to the bedroom.

Sources seem to vary. Some limit their definition to speech, while others include thoughts or feelings. I can see a valid argument either way.

"Ordered" implies speech, even if none was quoted, but surely "He sent them to bed" would not be considered implied speech, as there is no speech at all. That more similar to the other situations where you're reporting a fact or opinion.

It was a disaster.

That dress is ugly.

While these are thoughts, I do not feel that they belong in the same category as the others - even if you argue that indirect speech includes thoughts or feelings. They may be summarizing something someone said or felt, but there is nothing even remotely resembling a speech construct going on here, and the grammatical patterns do not apply.

  • +1, yes, it seems that for it to be considered as reported speech, it probably ought to have a reporting type of verb in the matrix clause or attached to it (e.g. say, tell, wonder, think, admit, beg, warn, suggest, remark, inquire, etc.). :)
    – F.E.
    Commented Nov 6, 2014 at 0:38
  • @F.E. So what's your opinion on the examples in the original question? Commented Nov 6, 2014 at 12:06
  • I upvoted your answer, btw, but isn't this really an answer to that other question that got closed? ;) Commented Nov 6, 2014 at 12:53
  • @Araucaria - I think it answers both? As I mentioned in my comments to the OP of the other question, I thought that there was interesting fodder for discussion there, but asking a question as checklist of which items were/were not reported speech was not showing the degree of research and thoughtfulness expected in a good question. Yours did!
    – Lynn
    Commented Nov 6, 2014 at 15:03
  • @Araucaria Both examples use reporting verbs, so I don't see the problem.
    – F.E.
    Commented Nov 6, 2014 at 16:49

It is pretty much a literal interpretation of 'reported' and 'speech', if you accept 'speech' as including written ('her email said...') and internal speech because you have access to somebody elses thoughts.

Turning to your examples, they are not reported speech in themselves. They are questions to find out if a particular item of speech was ever uttered. Essentially: "Can you report that blah blah was said?"

What I do find strange is that in both of those examples, I would say "I would be invited", not "I will be invited"

  • Whilst I don't necessarily disagree with your assessment ( - or agree either) do you have anything weighty to back up your claim that that is the accepted definition, seeing as we're already making exceptions for both writing and thought? And seeing that CaGEL, a pretty hefty authority seem to disagree on the basis of their presentation of these examples? Btw, thanks for your contribution! :) Commented Nov 5, 2014 at 13:41
  • 1
    Darn - deleted my own comment trying to hit edit! To repeat then... doesn't your own quote from CGEL cover it? That confirms that writing and unspoken thoughts are covered by speech. I wouldn't call them 'exceptions' as we often use phrases such as 'the instructions say' or 'her email says' for written communication, and 'internal speech' is a common enough phrase and is heard by ourselves, or by the omniscient narrator in a work of fction. What makes you think that CGEL is saying that questions about reported speech are reported speech? Commented Nov 6, 2014 at 11:54
  • Well, just that they have a section on Indirect reported speech p. 1024. It starts off with "The two main constructions used for indirect reported speech are illustrated below". They then give four examples, including the two questions in my examples. That would seem to be pretty conclusive evidence that they consider them to be reported speech! ;) Commented Nov 6, 2014 at 12:02
  • 1
    I am afraid I don't have a copy of CEGL, and I am reluctant to dismiss it without looking at one. However, my Swann English Language and Usage does say, in section 481 on page 503, that the auxilairy do is not used in reported questions (except negatives - "she asked why he didn't work harder") which does support my position that your examples are not reported questions in themselves, but questions about reported speech. The other clue is to ask yourself what the direct speech they are reporting would be. Commented Nov 6, 2014 at 12:41
  • Yes, Swann's quite right about that, but he's talking about the subordinate clause, not the main clause. So, for example Did she say whether he swims? is ok, but Did she say does he swim? is not normal ... As for your other points - they're good points! :) Commented Nov 6, 2014 at 12:47

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.