Does Anyone Know How We Did That? : The Evolution of the Periphrastic Do


Jason Aho


Copyright 2001




The verb do is a hard-worked word in present day English.  It sees uses not only as both a transitive and intransitive verb, but also as an auxiliary particle in many English language conventions. As an auxiliary, it functions as an emphatic particle (Do be a good boy John.), a yes or no question marker (Don’t you like peanut butter?), and a negative statement marker (I don’t know how it happened.).  When do occurs in examples like this, it contains no meaning on its own; this is called the periphrastic use of do.  Do was not always this burdened with work, however.  The following investigates the development of the periphrastic do, and is closely based on Olga Fischer’s excellent chapter in the Cambridge History of English Language.


Uses of Don in Anglo-Saxon 


When we look back towards do’s Anglo-Saxon roots, its resume reads very differently.  At this stage in English development, the OE don was PDE do’s ancestor.  Don was primarily used as a full lexical verb as in this example drawn from the Genesis story:


Uton . . . don hyne on þone ealdan pytt

         ‘Let us do [=put] him in this old well’ (in Fischer, 267)


 but sometimes was also used as a stylistic substitution for another full lexical verb in a clause.  These substitutions are known as the “anticipative” do and the “vicarious” do.  Anticipative do precedes the full lexical verb as in:


         utan don swa us mycel þearf is, habban     afire rihtne geleafan

         ‘let us do what is necessary for us, (i.e.) to have the true faith (in Fischer, 268)


whereas, the vicarious do followed the full lexical verb:


         …he miccle ma on his deaðe acwealde, ðonee he ær cucu dyde

         ‘he killed many more in death than he did before’. (in Fischer, 268)


OE also saw don used as a causative verb used to illustrate agency, as in:


         Đis hali mihte ðe dieð ilieuen ðat . . .

         ‘This holy virtue which causes one to believe that . . .’ (in Denison, 257).



Development of the Periphrastic


There are a large number of theories that identify causes for the introduction of the periphrastic do.  Like all inductive endeavours, however, there is no clear path to the truth.  Likely, there is a little bit of truth in almost all of the various explanations that have been put forth.  The most general explanation that can be offered is, of course, related to the progressive loss of inflection in English.  It only makes sense that when one method of conveying meaning disappears, other methods must be introduced. 

         It is only in the Middle English period that the periphrastic use of do comes into common practice.  Linguists have provided a variety of reasons for this development, saying that the emphatic use of do sprang up from borrowing from the Celtic or French tongues.  But, as Fisher makes clear, these theories are “generally ruled out” (269).  It has also been suggested that the periphrastic developed out of one of the substitutive uses of OE don, but this theory is unlikely since substitutive examples of don are not usually followed by an infinitive, and there are examples of periphrastic do that predate unambiguous substitutive examples that could be periphrastic.  

Olga Fischer argues that it might be that do has an inherent propensity to “develop into a semantically empty verb” (270).  Do, perhaps because of its general and multitudinous meanings, is a likely candidate to be transformed into an auxiliary verb, especially when the language requires syntactic clarification, such as could have been the case in English’s already well patterned system of auxiliary verbs. 


The Likely Suspect: Causative Do


It seems likely that the periphrastic use of do developed out of the causative use of do. Alvar Ellegård proposes that there were two related types of do constructions: one contained a noun phrase that functions as the object of do and the subject of an  infinitive as in: “Þe king dede þe mayden arise” (in Fischer 271).  This sentence is undeniably causative and means “the king caused the maiden to rise,” but there was another construction that Ellegård calls do x which had no such noun phrase, and therefore had an ambiguous sense of agency.  Consider the sentence “He dude writes sende.” (in Fisher, 270).  This sentence is either causative and means “He caused someone to send letters,” or it is periphrastic, in which case dude is semantically empty, and the sentence simply means “He sent letters.”   

Ellegård illustrates that it was possible for a verb to express causative agency without the use of do however. Consider the statement “Henry . . . þe walles did doun felle, þe tours bette he doun” (in Fisher, 272); “did felle” and “bette” both seem to express causative agency.  If “bette” can be causative without do, then so could “felle”, and therefore, do may be semantically empty and periphrastic.  Denison supports this theory by illustrating the existence of similar equivocation in present day English.  The verb “bombed” is non-causative in the sentence “The pilot bombed Dresden,” but is causative in “Nixon bombed Cambodia” (in Fischer, 272).  With both modes of causative expression available, the stage was set for the use of do without “any causative implication” (Fischer, 272). 

  The ability to introduce a one syllable word that essentially contained no meaning was, of course, great help to the poets of ME, since do could be employed as an auxiliary to complete metrical rhythms in their poetry. 

The auxiliary do saw regular use in the translation of French texts where faire was used as an auxiliary.  The use of do as an auxiliary could also have been influenced by the copious number of analogues present in English, such as haten and leten.  As is often the case, words are adopted into the patterns of other similar words. 

This has been a brief account of the development of the periphrastic do, but it does leave many questions unanswered.  How does the use of do in negative statements, emphatic statements, and in yes/no questions resonate with the dominant theories?  Can these features be ignored without leaving too large a gap in the theory? 





For Further Reading.


Two other brief overviews of the periphrastic do are:


Lass, Roger.  The Shape of English.  London: J.M. Dent & Sons, 1987.


Smith, Jeremy.  An Historical Study of English.  New York: Routledge, 1996.




For an in-depth study of the development of the periphrastic do, the inspired reader may turn to the relevant chapters in the following two books.  Both are excellent gateways to further scholarship on the subject: 


Fischer, Olga.  “Syntax.”  The Cambridge History of the English Language. Vol 2. 1066-1426.  Ed. Norman Blake.  New York; Cambridge University Press, 1992.  207-408.


Denison, David.  English Historical Syntax: Verbal Constructions.  New York: Longman, 1993. 




And, we’re not the only folks doing this kind of stuff on the web; there is a very lucid article to be found at:


Zilm, Reudiger.  The Development of Periphrastic Do in English.  August 21 2001.