The following is from English Conditional Sentences in Wikipedia.

If I liked parties, I would attend more of them.

If it rained tomorrow, people would dance in the street.

The past tense (simple past or past progressive) of the condition clause is historically the past subjunctive. In modern English this is identical to the past indicative, except in the first and third persons singular of the verb be, where the indicative is was and the subjunctive were;

It implies that subjunctive mood is used in counter factual protasis in English. On the other hand, in French, imperfect indicative rather than imperfect subjunctive is used there.

1) How can this difference between English and French be explained?
2) How can the use of past verb forms for counter factual meaing be explained in French?

  • 4
    I'm afraid that any answer will just boil down to: "The difference is explained through the fact that English and French are different languages." Being different languages they are licensed to treat things differently. Jul 12, 2018 at 13:16
  • 1
    @jknappen, this question is about "why" and "how". It can be interesting from the point of historical linguistics and it would be useful for translators who work with both languages. I'd love to know the answer. Jul 12, 2018 at 20:30
  • 1
    @jknappen: If you don't take the question too literally, but rather as a "can you explain the history of this phaenomenon in both languages (and compare them)", then I think a good answer is possible, as, indeed, Michaelyus has given below.
    – Cerberus
    Jul 17, 2018 at 16:33

1 Answer 1


Essentially, regarding irrealis conditionals, it is French that has changed "further" than in English in this regard.

In Classical Latin, both the protasis (the if-bit) and the apodosis (the consequence) would both be subjunctive, either imperfect or pluperfect depending on tense.

Nam nisi multorum praeceptis multisque litteris mihi ab adulescentia suasissem, ... numquam me pro salute vestra in tot ac tantas dimicationes atque in hos profligatorum hominum cotidianos impetus obiecissem. (Cicero, Pro Archia)

For unless I had persuaded myself from my youth with the teachings of many, and by many books, ... then I should never, for the sake of your safety, have subjected myself to so many and such great struggles and to the daily attacks of desperate men.

But it seems Vulgar Latin / Common Romance went with what was the pluperfect subjunctive and had it replace the former imperfect subjunctive, and named it... the imperfect subjunctive. They then (mostly) dropped the subjunctive from the apodosis and replaced it with a new form, cobbled together from the infinitive with the imperfect indicative of (whatever became of) habeo ("have to", in its extended meaning of "must"), which became known as the "conditional" mood. [Note: Standard Italian ended up using the derivative of the Latin perfect indicative, which became the Italian passato remoto, usually called the "preterite" or "past historic". Old Tuscan did use the imperfect indicative of avere though.]

New compound tenses were based on this paradigm. Hence, modern Spanish (with pluperfect subjunctive and conditional perfect):

Si no lo hubiésemos explicado antes de la foto/ , no habría resultado sencillo saber qué imagen se correspondía con algo bueno para Neymar. (El País)

If we had not explained it before the photo, it would not have been easy to know which image corresponded with something good for Neymar.

Old French also did this:

Car la tenisse en France et Bertrans si i fusset, / A pis ed a martels sereit aconseüde. (Pélerinage)

As if I had kept it in France, and if Bertrand were there / With pikes and hammers (she) would have been attacked.

But this coexisted with the same construction as Classical Latin (subjunctive in both clauses):

Se je ne fusse en tel prison, bien achevaisse cest afere. (Le Vair Palefroi)

If I were not in such a prison, I would put an end to this matter.

... and with what Modern French would do, which would be with imperfect indicative with the conditional present.

Et se je pooie amander la mort don je n'ai rien mesfet, je l'amanderoie sanz plet. (Yvain)

And if I could pay death which I have never wronged, I would pay her without complaint.

By 11th century, the conditional tense was replacing the subjunctive in the apodosis, but the indicative replacing the subjunctive in the protasis took longer, persisting into early Classical French of the 17th century. One idea is that it naturally emerged in Late Latin where the Latin imperfect subjunctive was disappearing and hence they needed something for this present counterfactual construction, with the fact that it was more "present" being allocated to the imperfect indicative.

However, as one forum discussion explains, the disappearance of the imperfect subjunctive from the protasis is the major thing that needs explaining. One article from 1901 suggests that the imperfect subjunctive fell out of favour because the endings -asse/-asses/-assent (all pronounced /as/ in IPA) ended up sounding "coarse" because of its relation to the derivational suffix -asse (those who know their French swearwords will testify to this). Interesting theory, but not necessarily satisfactory.

What happened then to the past subjunctive in English? It seems to be me that it was swept up in the general reduction of endings, and then the replacing of constructions using the past subjunctive conjugation with auxiliary verbs like should and would. The bulk of this change happens in Middle English. But the past subjunctive persists very clearly in Middle English, as Middle French translations from the period attest.

However, Old English does not seem to use the subjunctive forms in conditionals, and it is Middle English that appears to innovate this (in both factual and counterfactual), with past subjunctive in both protasis and apodosis in counterfactual conditions, although normally the apodosis uses a modal auxiliary in past subjunctive (e.g. sholde). Old English does use subjunctive forms a lot more as a whole though.

and I were a pope/ Not oonly thou, but every myghty man,/ … Sholde have a wyf;

If I were a pope, not only you but every mighty man would have a wife.

Another factor that seems to keep the subjunctive in Middle English is the use of inversion to signal the protasis, instead of explicit conjunction if, unless. This persists into modern English, in relatively set phrases:

Were he still here, ...

In conditional sentences, there is a moribund distinction drawn between indicative and subjunctive in English:

If he/she was singing, you would have certainly appreciated it.

If he/she were singing, you would have certainly appreciated it.

... relating to counterfactuality.

Is there any meaningful comparison to be drawn here? It doesn't seem like it to me - there is a completely different set of constructions that the two languages have charted to produce counterfactual conditional statements.

  • Excellent answer. Two notes. 1. Even in Cicero, it's easy to find examples of the indicative past tenses used in counterfactual protases. 2. I think the deeper reason why past tenses are often used to describe counterfactual types of sentences, as here, is that both a counterfactual situation and a past situation are different and removed from the present, real situation: the past tense creates some distance from the here and now.
    – Cerberus
    Jul 17, 2018 at 16:38
  • ... And in the protasis it's usually very easy to tell whether past or conterfactual was intended (almost always counterfactual), so using the imperfect indicative was clear and unambiguous enough.
    – Cerberus
    Jul 17, 2018 at 16:38
  • English and French replaced past subjunctive for different reasons, but the results are the same, past indicative. Is that purely coincidental? It seems surprising especially considering in the case of English it is the result of dropping the endings, which involves no semantics such as "remoteness".
    – Aki
    Jul 20, 2018 at 14:10

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.