Reported Speech and Conditional Sentences




As a consequence of this author’s library of textbooks of English completely lacking, or only summarily dealing with the question of conditional sentences in reported speech – a topic which had neither been dealt with in his years of study, nor in the courses he had taught at any institute – he deduced the treatment that such sentences should be given on the basis of the little available information. Upon further delving into the question, the conclusions were cross-checked with any additional findings.  While the intuitively-comprehended grammar was in agreement with those discoveries, it was felt that this material needed to be made more available to a wider audience which might desire to accurately reconstruct an original sentence from reported speech – a much more daunting task – and a question rarely dealt with.  Although the need for such a transformation might be rare, inquiring minds may come up with the questions dealt with in this paper, and it behooves the language professional to have a ready answer.


This brief article looks at an aspect of reported speech which is often overlooked in textbooks, perhaps because that facet should be considered as self-evident, and as a corollary of what was already learnt. Consideration of the correct treatment to be afforded in indirect speech is not lacking when speaking of words such as “say”, “tell”, “request”, “command”, etc.; and indeed, such a list of words could be appropriate when dealing with a sentence that was stated as a conditional. However, it would seem that to correctly convey the thought expressed in a conditional declaration, reported speech must take special pains to avoid confusion as to what was said verbatim in the original statement, and not approximately. A web search using the terms “reconstruction of direct speech from reported speech”  gave only one result among the first ten, and that was to a very specific question.  Should such a point of view seem to be of little practical use, there are nevertheless those students who might come up with the question dealt here, and it may well be worth avoiding a red face by having a ready answer. We herewith look at the causes and possible cures to such situations.

The reader is forewarned that some of the stranger sentences here reflect a quirky sense of humour. As a web-based article, attempts at gaining visible placement require some non-standard practices. From time to time, the criteria varies – but of one thing we can be sure: where we would like to give all possible transformations of a sample sentence, it was felt more prudent to leave that as an exercise to the reader.

Basics of Indirect Speech

As this writer has pointed out to his students, for all practical purposes, after a verb such as “said”, following the rule that the verb tense must shift back one level from what was originally used will achieve correct or acceptable results. Where it is possible to contemplate exceptions, it would do students well to consider the particular demands of the teacher, or of any test to be taken, so that one is not lulled into a false sense of security. Not all texts treat the subject of indirect speech with the same general solutions, and indeed, students who transfer from one school to another, let alone from one country to another, should not automatically assume that what they have learned as correct will be so considered in a different environment.

The following elementary principles reflect what this teacher considers as standard.

One need not know much about grammar to know that yesterday came before today, which is another way of saying that the past preceded the present. This truism allows us to deal with quotations in the simple present tense in reported speech, by changing the verb to the past. The same process is, of course, extended to the present perfect. (Other words in the sentence may require changing, but there is enough material to be found elsewhere on that topic.)

Speaker: “There IS enough food on the table of gluttons.”

Indirect Speech of the preceding: The speaker said (that) there was enough food on the table of gluttons. “IS” in the quoted sentence changed to “WAS” in oratio obliqua. (Treble and Vallens, 1936, 99.)

Here is an example with the present perfect, with emphasis on results affecting the present:

Health Authority speaking: “There has been an epidemic of overeating in the country during the last few years.

Indirect Speech of the preceding: The health authority said that there had been an epidemic of overeating in the country during the last few years.

The English language is not so kind in providing us with suitable words for describing something which happened in a time past before another time gone by. Spanish has a word, antepasado, literally meaning “before (the) past”, but the first dictionary we looked at gives us the unsatisfactory renditions “passed” and “elapsed”. Another gives us “before past”, but using that in a sentence related to grammar might provide a minor challenge. In light of the next Spanish word we present, it would seem strange that this seemingly “before-the-past” has nothing to do with verb tenses, but with ancestors.  We bring it up only because it seems to be a “missing link” in a series of terms related to time.

Of course, we have the term “past perfect” or “pluperfect”, which are more arcane than the Spanish term we have given. To keep things simple, we will point out that one Spanish translation of the present perfect is antepresente. As in English, though, different authors may vary in their usage of terms, and, for example, the list of past tenses and their alternative names in (Bello and Cuervo, 1945, p. 164) is formidable. However, in neither English nor Spanish do we have a solution for giving a rendition of time even more previous to a such a period already defined prior to another. The following diagram shows that Spanish did make a minor concession to a time period immediately preceding another, but that is as far as it goes. We must, of course, not fall into the trap of thinking that “before-the-present” implies the (simple) past, any more than to attribute a greater completion of anything in the here-and-now by the present perfect.

Past perfect —- simple past –Spanish antepresente — present perfect — simple present

If the present perfect were taken as a non-grammatical term, and understood literally, who would use it? We may postulate the Biblical characters, Adam and Eve. One says to the other: “We are living in the Garden of Paradise, in Eden, in the present perfect, in a perfect present.” Of course, they were yet to learn that the present does not always have such a property. And Bible readers can tell us, for this couple, the grass was not greener on the other side, that is, it definitely was not a future perfect!

Teachers might consider asking students how to make the following sentence sound as if it were “more in the past” than it is at present; individual readers will find the solution below:

In a sentence with the present perfect, e.g.,

Speaker: “There HAVE BEEN enough wars in recorded history”, it is not even necessary to know the name of the tense arrived at (if results are to be merely intuitive), because the present tense of the auxiliary verb changes into the past again, giving:

The speaker said (that) there HAD BEEN enough wars in recorded history.

Because we see no end to wars, the preceding sentence might easily be up for an examination as to whether the tense change is required, but we will follow the safe rule.

The student who needs to name the transformation has no choice but to know that we have now used the past perfect, a.k.a. the pluperfect.

While a sentence has already been given in the present perfect in direct speech further above, we prefer, in that example, to think that the epidemic referred to is still with us in the present; while in this example about war, we would like to think that the problem is close to the present, but fading away. That such a sentiment is wishful thinking is another matter.

The teacher might have students try to put the following Lewis Carroll type nonsensical sentence into reported speech, to see what thoughts come to them:

Moderator: Before the able fashioning of a cane, when wood WAS scarce, there HAD BEEN a dam on the river on the eve of creation.

If one were to suppose that our grammar is lacking in the required tools to put this sentence into a form even more in the past – something that seems to be rarely discussed, we can reply that the event before another makes time relations clear, even in reported speech:

The moderator said that before the able fashioning of a cane, when wood WAS scarce, there HAD BEEN a dam on the river, etc.

Hint: To better understand the sentence, try reading it as “There had been …., before ….”

Reported Speech with Conditional Statements

Based on the foregoing, there should be, at first glance, no problem with the reported speech of a sentence like:

Science Teacher: If a scuba diver comes up too fast, [s]he will get the bends.

The above, a zero conditional, or type 0, involving statement of a fact, with the “if” meaning “whenever”, is not a conditional at all. Reported speech in this case will not have any tense change for the words that were spoken.

The teacher explained that [if / when / whenever] a scuba diver comes up too fast, [s]he will get the bends.

Nor, by following standard rules, there will not be any problem with the first conditional, the real conditional in the present, nor in any other case where the main clause is in the present tense:

Disenfranchised Person: If I GET the vote, I WILL elect someone who IS in favour of justice.

Our reporter declared that the disenfranchised person WOULD elect someone who WAS in favour of justice if [s]he GOT the vote.

There is no problem here, but, if the first conditional – the real conditional in the present – is in effect the same as the second conditional, a difficulty in interpretation could arise. Let us see what I can invent:

Note: Search engines consider repetition of key phrases as a sign of spamming or unfair attempts to direct traffic to one’s page; so, while desiring to change only the verb tense in the following examples, it has been necessary to revise each sentence to something sufficiently different. The reader’s indulgence and imagination is requested. The italicized segment is what was originally retained in all sentences of a similar nature. We have presented the desired text in image format, to try to elude the problem mentioned.


Indirect Speech/Conditionals

Complete set of conditional sentences and the resulting reported speech. Click image to enlarge.

If I forget my solid-gold watch in the washroom of a greasy-spoon restaurant, it will not be there a couple of hours later.

Since this is a simulation of a conversation, let us ignore, in this discussion, the separation of the word “it” from its proper antecedent. The standard reaction is always, “You know what I mean!”.

The preceding does not change much in meaning from the next, taken as a hypothetical conditional:

If I forgot my silver bracelet in the lavatory of a ghetto greasy-spoon, it would not be there after a two revolutions of the minute hand of my watch.

Note that if the first sentence were put into reported speech, its completion would be in the form of the second conditional, i.e.:

The debutante said that if she absent-mindedly neglected to pick up her jewels from the vanity of the ladies’ room of a common diner, they would not lie there anymore after she had finished her four-course dinner.

Reported Speech:

The writer said that if he had forgotten his solid-gold watch in the washroom of a greasy spoon restaurant, it would have got[ten] stolen.

How could that be distinguished from the treatment to an unreal conditional in the past (third conditional) such as:

“If I had forgotten my solid-gold watch in the washroom of a greasy-spoon restaurant, it would have got[ten] stolen”?

Since we cannot change the past perfect in the unreal conditional, the reported speech of the 3rd conditional seems to be equivalent to that of the second form – which just might mean the same as a first type, albeit with a slight nuance about the probability of the event taking place. We have looked to see how textbooks treat such a situation, and originally came up with a blank.

Solutions to Ambiguities in Reported Speech of Conditional Sentences

At first sight, at least for the individual with no other data to work on, the conditional sentences would be treated as ordinary sentences, and the reported speech would follow the rules that the student already knows.

The standard treatment according to this writer’s library is to change the tenses of the first conditional, but not in the second and third conditional forms. [Alexander, 1988, p. 292; Thomson and Martinet, 1980, pp. 195-6, or 1986, p. 205]. B. D. Graiver (1986, p. 103) gives reasons for this, but does not follow the usual “[s]he said” routine in the examples of reported speech. It is surprising to note that Quirk, et. al. in A Grammar of Contemporary English do not have any references to this, while Quirk, et. al., in A Comprehensive Grammar of the English Language, (Longman, 1985), only give a fleeting reference to the “past subjunctive” or “hypothetical past” being backshifted to the “hypothetical past perfective” (p. 1031).

Since our question was posited and pondered before recourse was made to reference books, the conundrum was solved by treating sentence of the nature described above just as they are – not as mere declarations, but as something either ascertained (1st conditional) or hypothesized (2nd and 3rd conditionals).

For the real conditional, we might therefore, to avoid any ambiguity, declare that:

The writer stated WITH ALL CERTAINTY that if he left his solid-gold watch in the washroom of a greasy-spoon restaurant, it would get stolen.

Our words, “with all certainty”, or anything to that effect, would show that the conditional sentence which had been used was made with the present and future tenses. We could, more succinctly, replace “stated with all certainty” by “averred”, assuming that it will be understood as “stated positively”.

Upon further contemplation, it seems that our solution is too much in the line of what was found for reported speech in Collins Cobuild Student’s Grammar (1991, 92). We have deviated a bit too much from “said” – not as far as the just named work – but we leave open the possibility that we incorrectly think the original sentence to have been something like: “With all certainty I left …, etc.” So, while this option is available, this author would prefer an improvement, something along these lines:

The diner stated, with what might be considered/what one might consider a high degree of certainty, that … etc.

This way, it is clear that the person who is talking about “certainty” is the reporter, and not the individual under discussion.

For the hypothetical conditional in the present, with a meaning almost equivalent to the above:

The writer hypothesized, with a high degree of certainty, that if he had left his solid-gold watch in the washroom of a greasy-spoon restaurant, it would get stolen.

A potential substitute is “supposed”, “surmised”, or perhaps, “guessed”. Again, we could criticize the expression with “certainty” as ambiguous. Substitute words in this paragraph are the same or similar to those in Collins. A word-for-word comparison has not been made.

Finally, to distinguish the preceding from a totally unreal expectation:

The writer concluded that, in the event that he had been so careless as to leave his solid-gold watch in the washroom of a greasy-spoon restaurant, it would have got[ten] stolen.

Grammatically, there is nothing different from the books that were consulted – it is our choice of words about the way the report was delivered which is different.

By changing our “if” to “in the event that”, we suggest that the action did not take place. For even more clarity, we might say “in the unlikely event that”. A subjunctive, or its equivalent, would also be possible:

Rabelais concluded by saying that, had he been so careless as to leave … .

This solution allows for a second conditional treatment in such cases where it is not intended to convey the idea of a lesser probable first conditional:

Molière supposed that if he were to have left … .

This writer expects that the number of times anyone would have to manage constructions such as those illustrated above are rare, but if he had to ask himself how to make the necessary transformations, others might have asked themselves the same question.

These solutions are especially recommended to people reporting the news and who would like to earn a reputation for both honesty and precision in their reports.

November 4, 2017, updated June 2, 2018.
© 2017, Paul Karl Moeller

Sources Consulted

Note: The web source omits the prefix https:://www. to avoid linking. Some browsers will work with the site’s name as given below; otherwise, if the link still exists, add the missing prefix.

Alexander, L. G. , Longman English Grammar, 1988.

Bello, Andrès, and Cuervo, Rufino J. “Significado fundamental de los tiempos compuestos del indicativo”, Gramática de la lengua castellana, Buenos Aires: Ediciones Anaconda, 1945).

Gayathri Narayan, “How do we change an indirect speech with reported verb in past perfect into direct speech?”,

Graiver, B. D., Advanced English Practice, 3rd ed., Oxford, 1986.

Thomson, A. J., and Martinet, A. V., A Practical English Grammar, 3rd Ed., Oxford, 1980

Thomson, A. J., and Martinet, A. V., A Practical English Grammar, 4th ed., Oxford, 1986.

Treble, H. A., and Vallins, G. H. “Indirect Speech”, An ABC of English Usage, London: Oxford, 1936,

Quirk, Randolph, Greenbaum, Sidney, et. al., A Grammar of Contemporary English, Longman, 1972.

Quirk, Randolph, Greenbaum, Sidney, et. al., A Comprehensive Grammar of the English Language, Longman,. 1985.

University of Birmingham and Collins Cobuild, Collins Cobuild Student’s Grammar, HarperCollins, 1991.


%d bloggers like this: