Voices of Earlier English: Robert Robinson reads Shakespeare

In 1617 one Robert Robinson, a kid who matriculated pensioner from Trinity Hall two years before, composed one of the most original works of early English phonetics in "The Art of Pronuntiation" in which he used a script deliberately unrelated to the Roman alphabet to sidestep its limitations. He lived as a schoolmaster in London and apparently fourty years later impressed Charles Hoole (teacher of a grammar school in Lothbury Garden) with the speed with which he taught young children to read. Hoole implies that Robinson was using a technique peculiar to him. He appears to have been poor, but had a good education.

His published essay is a general theory of phonetics, complete with a series of symbols by which he hopes to be able to represent speech (of any language) unambiguously. Essentially, Robinson was trying to create a kind of Early Modern IPA. It is not without problems. His desire for featural elegance led him to express things like tenseness in a roundabout way. In this he was probably influenced not by traditional spelling as is commonly suggested but by the model set by Latin and Greek grammarians. His description of the distinctive features is very original and sometimes quite perceptive. I actually like his category of "lesser obstrict" a lot. He seems to have independently arrived at the rudiments of ideas like place and manner of articulation.

He then used his proto-IPA to transcribe his pronunciation of English and Latin. And in doing so is able to unproblematically represent a few phones that anyone bound to just the Roman alphabet would have a hard time doing. He reports things like optional allophonic devoicing, the fact that his particular dialect had a realization of /kl/ and /gl/ clusters that we would otherwise not know existed in the city of london at that time.

In his deliberate disregard of the Roman alphabet, he allows himself to recognize variety not just among different people but in his own speech. He feels no obligation to transcribe every instance of a word the same way if he doesn't always pronounce it the same way. Unlike Hart and Gil and all the English spelling-centered sources, the phenomenon of free variation causes Robinson no ideological problems. He is the only such witness to proceed in such a descriptivist spirit. His idiolectal transcriptions give us real information about an incomplete merger of *ā and *ai in his idiolect, one which was only partially lexically restricted.

His book did not sell well, and only one copy of it survives.

His proto-IPA is rather clunky and in some ways fails him. His biggest limitation is that he seems never to have really learned to speak any other language. Had he broadened his articulatory experience, who knows what he might have accomplished. In any case it's a shame nobody in his day realized that he was onto something big in featural representation. Despite some of its awkwardness, his script does allow one to build up a fairly well-supported sense of what a young well-educated but clearly sub-elite man might sound like when reading Shakespeare's sonnets which had been published just eight years earlier. Especially since he transcribes his pronunciation of verse-texts in his private papers.

The pronunciation Robinson records himself as using in reading verse is a good deal more...formal than the one I typically use for these readings. Much more sparing in its use of weak forms of function words. He usually gives words like "my" or "I" or  "have" the vocalism associated with the strong forms. And when reading verse he always pronounces the <l> in "should/could/would". Incidentally, this is one guy for whom their/they're were NOT homophones. So I read it in a careful way, quite performatively flat, the way a schoolmaster might when trying to illustrate pronunciation.

So here you go, sonnet 18 in his English.

(If you're wondering why this doesn't sound like David Crystal's "OP", here's a tediously detailed disussion of everything that is wrong with it.)

Historians of English phonology haven't always recognized Robinson's achievement fully. Chomsky and Halle were among the worst offenders. Some of what Robinson has to say conflicts with certain theoretical presuppositions about the way in which important sound changes took place. The thought has often been that Robinson (and Hart, and Gil, and pretty much everybody else) must have been misled by English spelling or failing to properly perceive things that "ought" to be there.

People like Dobson (Robinson's most recent editor), Kökeritz and Cercignani never forgave Robinson for not representing the PRICE and MOUTH vowels as sharing a single centralized onset. Worse still, there is really is no way to make sense of how he sets up his feature grid unless ME /ā/ was not yet /ɛ:/ for him. Otherwise he could put it in a length-pair with ME /ĕ/, while pairing ME /ă/ with pre-rhotic ME /ā/. The very fact that he reports a distinction between ME /ā/ and /ɛ̄/ in pre-rhotic contexts and equates the former with ME /ā/ in free development simply makes an [ɛ:] value for ME /ā/ unlikely unless there is something very peculiar about his rhotic. But a value [æ:] would mean that ME /ā/ and /ai/ were merging to [æ:]. As Tessio said in the Godfather "hell, he can't do that; that screws up all my arrangements". Dobson for his part found it preferable to think that /ai/ could only ever monophthongize one way in the history of English. He had his reasons. Philologists dealing with the history of a national language have sometimes suffered from morbidly severe cases of teleology.

Wolfe and Lass set the point straight: a man like Robinson who could describe his feature grid this way, who could tell that orthographic <cl> in <declines> actually corresponded to [t͡ɬ] in his pronunciation (a feature now restricted to certain moribund dialects outside London), who was so dissatisfied with English spelling that he came up with an entirely new script, was not being blinkered by naive reliance on traditional orthography in his transcriptions. The only influence of orthography, I think, is that sometimes variant spellings elicited different covarying pronunciations when he read text aloud.


  1. Utterly fascinating and thought-provoking

  2. Simon Roper brought me here and I'm glad he did. Fascinating.

  3. Very interesting! I'd love to hear more read in this accent.

  4. What a wonderful discovery! Not just Robertson, but your blog in general.

  5. I came here from Simon Roper, too. This is a very nice effort, one of the most interesting reconstruction attempts I have heard!

  6. Dear mr. Foreman,

    In an older blog (http://blogicarian.blogspot.com/2019/03/argumentum-ad-ignorantiam.html) you mentioned a Latin speech (https://youtu.be/a61Dc_EFuI4) by Luigi Miraglia. You wrote that some untrue (or half-true) things were said in it. I would love to know more about the things he said which you think were missing the mark!

    Warm regards from Amsterdam,

  7. If true, this is fantastic. But I need to know more: which "moribund dialects outside London"? You can email me at connie_scozzaro@brown.edu

  8. One interesting fact is that this reading is completely understandable to me (a 79 year old American) even though I have never read or heard this sonnet before.

  9. anything that disagrees with Crystal is fine by me :-)

  10. I have no words to describe how precious this is1 I already followed your other blog Poems Lost in Translation and am happy to have decided to check this one. Thank you so much for your amazing blogs, they're cultural treasures themselves.

  11. This, agrees very well with the "Shakespeare Original Pronunciation" going around, with distinctions well within the range of dialect.

  12. sounds like a north Dublin accent to me!