We are familiar with the concept of "do support", where the verb do is used as an auxiliary verb. It can be found frequently in Shakespeare and before and it is claimed to derive from the Celtic languages of Britain. It is suggested that a different verb with the same meaning was used in the Brythonic (Welsh-like) languages, most commonly in declarative lists such as "jumping, shouting and screaming did he".

But neither this Brythonic use nor the Shakespearean use match the most significant feature of do support in Modern English, namely that we have to use it in most questions and negatives.

So where does the rule saying that we have to use it in questions and negatives come from? I can't find anything anywhere.

  • 1
    Any answer should address the variety of do-support (Brythonic, Shakespearean, emphatic, compulsory, questions, negation, etc etc) along with example sentences so that we know which one you're talking about. I think the most likely answer is that any description of 'do support', whatever its provenance (Celtic vs Germanic vs Late Middle English), covers all the different varieties, including question/negation auxiliary 'do'.
    – Mitch
    Feb 3, 2022 at 22:10
  • 1
    See duplicate of: english.stackexchange.com/a/218364/18250
    – Alan Munn
    May 5, 2022 at 4:53
  • There isn't a rule saying that. There are simply more syntactic rules in Modern English that require auxiliary verbs. Negation and question formation are the most common, but there's also contractions and tags, which are intermixed, and pro-verbs like I ate more than he did. Auxiliaries are much more important to a language with few paradigms than endings ever were. And do has gotten obligingly bleached of its active meaning (see Ross's "Do" paper) and can stick in for anything, even stative verbs. It's there to seal off the beginning of the verb chain, like modals. Sep 4, 2022 at 16:03
  • Actually some dialects (probably by now either moribund or defunct) do actually use 'do' in the way questioned. Some 50 years ago when living in Somerset, I would hear some people still saying things like "I do be going into town this afternoon."; "We do start eating every day at six." Similarly some people were still saying "I be, you be, he be, we, be, you be, they be." So the use of 'Do' described in the question need not be a strange as all that.
    – Tuffy
    Dec 27, 2022 at 19:39
  • 1

3 Answers 3


The fossilised, original "do support" is in fact common to all modern Germanic languages - it's unlikely there's any syntactic Celtic substrate in English - documented all the way back to Gothic and probably originating in Proto-Germanic; with the past tense of weak verbs formed with the infinitive stem suffixed with the past tense of the strong verb "to do" (I "walk-did" = "I walk-ed"). That also accounts for the earlier conjugations.

Extant Old English texts misleadingly only present a highly conservative/prestigious form of the language, likely diverging from the spoken language in terms of syntax and semantics as much as Latin did from the Proto-Romance. Let's imagine that among the illiterate masses - that's almost the entire population - this "walk-did" in interrogative sentences was gradually re-analysed from "Walk-did you to the village?" to "did you walk to the village?". Plausibly, this then found its way into present-tense questions, which then saw "do" being mirrored in the response: "Do you like apples? Yes, I do". That in turn gave us the emphatic (slightly archaic?) "Oh I do wish you'd stop it".

It's worth noting that the "do support" seems to have flourished in the post-Norman Conquest period and beyond, when written Old English - bearing scant resemblance to Norse-influenced, increasingly Frenchified Middle English - was abandoned for good. Perhaps also relevant is the fact English - and Afrikaans? - is the only Germanic language lacking nuance-adding modal particles (Ich liebe dich doch! = I do love you!), with the "do support" also making up for this to a certain extent.


There is an interesting paper "The rise of auxiliary DO: verb-non-raising or category-strengthening?" Richard Hudson UCL https://dickhudson.com/wp-content/uploads/2013/07/histaux.pdf

The paper also considers developments in the auxiliary system since the 16th century and offers a mixture of cognitive and functional explanations for the changes since the 13th century.

Auxiliary DO is often called `periphrastic' DO because it has no meaning independent of the meaning of the construction concerned; the only reason for using auxiliary DO in Modern English is because the syntax requires an auxiliary and no other auxiliary is needed by the sentence's meaning. DO fills the gaps where non-auxiliary verbs are not allowed and where other auxiliaries are not needed.

The Brythonic hypothesis is probably wrong as it has had virtually no influence on English. From the same paper:

The historical facts relevant to negative and interrogative sentences can be summarised as follows (I follow the account in Denison (1993:255f)). The `periphrastic' (semantically empty) DO appeared first in the 13th century, but at that time it had no special connections with negation or inversion; nor, indeed, were these restricted by rule to auxiliary verbs.

The 13th century is a little late for any Welsh influence.

In addition, from Old English Online we have

Finally, it is worth mentioning that although the verb 'to do' was used regularly in Old English, it was not used as an auxiliary in the formation of questions or negations as it is in modern English. So we would find 'He swa dyde - he did so', but 'Dranc he? rather than 'did he drink?' and 'ne dranc he' rather than 'he didn't drink'.


It's not a rule per se, so much as something everyone does. Throughout the history of the English language, if the correct meaning is still conveyed, grammar rules tend to evolve with the current usage.

To create a question in English, you invert the subject and auxiliary/copula word of a declarative sentence. Making a question out of the following sentence is easy:

I will know.

Will I know?

It's easy to see that will and I should trade places. But how do you make this sentence into a question:

I know.

This is a vaild sentence in English, yet it has no auxiliary/copula word. There is nothing to trade places with I.

When sentences have auxiliary/copula words, the rule is easy to apply, and the resulting sentence is not awkward. Strict adherence to the rule in the absence of an auxiliary/copula word makes for an awkward sentence:

Know I?

If you use a proxy verb in place of a missing auxiliary/copula word, then you don't need the rule to ask a question. do is a great candidate for a proxy verb, as it is a Swiss-Army knife for the English language.

Do I know?

I'm betting printing press operators helped perpetuate and encourage the use of do in these cases, as they did not have to rearrange letters within words to form a proper question. The more people read the sentence with do, the more they used it as well.

Consider more how English is used, rather than what the rules are. Today's rules are defined by how English was used in the past, and tomorrow's rules will be defined by how it is used today. Rules try to nail down how the language is used, but they don't always reflect the lay person's actual use. do-support is part of the language now, but I'm betting it evolved organically rather than appeared suddenly.

  • 2
    Could you clarify your argument involving printers? To set the text of a question in type, a printer need only arrange the letters of that text itself. They needn't take anything different and rearrange anything.
    – Rosie F
    May 5, 2022 at 7:37
  • 2
    I'm sorry, but this is simply wrong on all counts. Please see the answer linked in the comment to the question.
    – Alan Munn
    May 5, 2022 at 14:05
  • 1
    I don't get your argument about what is wrong with > Know I? which was the original point of the my question. The French do it. The Germans do it. Shakespeare did it (interchangeable with the do construction). In the Celtic languages, VSO is the normal word for a statement, with additional markers but no change in word order for questions and negatives. The only reason it sounds odd in English is because we don't do it. The argument becomes completely circular if you then argue that we don't do it because it sounds odd. May 8, 2022 at 21:57

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.