P is for (Thomas) Prendergast

6 08 2017

Thomas Prendergast.jpgThe mention of Thomas Prendergast in my last post sparked a couple of enquiries. Who was he and what was his method?

For all his working life, Thomas Prendergast (1806 – 1886) was, like his father before him, a civil servant in the East India Company, during which time he learned at least two of India’s indigenous languages, Hindustani and Telugu. On retirement in his fifties, he returned to England where (now blind) he spent his remaining years developing what he called his ‘Mastery’ system, published in 1864 as The Mastery of Languages or, the Art of Speaking Foreign Tongues Idiomatically, along with accompanying teaching materials in a variety of languages.

Essentially, his method involved the cumulative memorization of a set of exemplary sentences and the subset of sentences that can be derived (or ‘evolved’) from each of them. With remarkable foresight, Prendergast had observed that children seem to achieve fluency by memorizing entire sequences of words – what we would now call chunks or constructions. They are also able to combine and re-combine these elements in creative ways. Accordingly, Prendergast set about trying to identify the most frequent constructions, albeit only those that qualified as well-formed sentences. Prendergast’s sentences were not graded from the simple to the more complex. Rather, they were deliberately contrived to pack in as much syntax as possible, the test of their usefulness being the number of less complex sentences that could be generated from them. Because children are able to derive the grammar from constructions without explicit instruction, Prendergast was adamant that all grammar explanation was ‘prohibited’.

Prendergast’s Mastery system seems to have enjoyed some degree of success in its time and was adapted to the teaching of a number of languages. It was soon overtaken, however, by the arrival of the Reform Movement, and the kind of ‘direct method’ that was popularized by M. Berlitz.

Nevertheless, in many ways, Prendergast’s system prefigured developments in methodology that were way ahead of their time. One of these was the use of what later came to be known as substitution tables: i.e. tables that display the way that words and sentence elements can be combined. Also, his belief that mastery of a limited ‘core’ of structures and vocabulary could serve as a foundation for later proficiency contrasted with his contemporaries, for whom principles of selection or grading were largely ignored. But perhaps most remarkable was his insight that fluency, at least in part, results from having a memorized store of fixed and semi-fixed formulaic utterances. Unfortunately, by supposing that these ‘chunks’ consisted of whole, syntactically well-formed sentences, his method – in Howatt’s words – ‘turned the wrong corner’ (2004, p. 176). It would take another half-century before this misstep would be corrected, and the units of fluent production would be re-envisaged, neither as words nor sentences, but as ‘word groups’ (Palmer 1921).

Nevertheless, to give you a flavour of just how innovative Prendergast was, here is a selection of quotes:

On grammar

‘Although no one has ventured to maintain that the words “language” and “grammar” are synonymous, there prevails the notion that a knowledge of grammar is equivalent to a knowledge of the language to which it relates’ (1868, pp. 78-79).

‘Grammar is sometimes defined to be the law by which language is regulated; but in reality, grammar is deduced from language, and is not the regulator, but the regulatee’ (1864, p. 191).

‘No definition of the term “grammar” enables us to understand why that science should be studied first’ (1868 p.65).

‘The definition which styles it “the art of speaking correctly” has so little truth in it, that many persons who are well-versed in grammar are either incapable of speaking at all, or else, when compelled, are so embarrassed by the conflicting recollections of rules, exceptions, cases, tenses, moods, and genders, that they cannot help speaking incorrectly. The grammar itself is the cause of their speaking ungrammatically’ (1868, p.79).

‘Usage is the only law. Usage constitutes the whole code’ (1864, p. 203).

 Mastery title pageOn acquisition

Studying a language is not acquiring it’ (1864, p. 200).

‘Some say that we must think in a foreign language before we can speak it well … But it is not by thinking in a language, but by not thinking in it, that children speak it idiomatically and fluently’ (1868, pp. 233-4).

‘Illiterate people and children acquire the power of speaking the most difficult languages with fluency, by learning a very few practical sentences, and by ringing the changes on them’ (1864, p. 209).

‘Children and imbeciles succeed, in spite of their ignorance of grammar and books’ (1864, p. 209).

 On vocabulary learning and chunks:

‘[Oral] composition is not the compounding of sentences according to the prescriptions of the grammarian; but it is the putting together of idiomatic phrases by intelligent efforts of memory’ (1868, p.32).

‘[Children] import an idiomatic combination of words, together with the ideas belonging to it; they immediately begin to employ it for practical purposes without alteration; and they repeat it so often that it becomes stereotyped in the memory’ (1864, p.34).

‘Language is a tree which is propagated not by seeds, but by cuttings; not by words but by sentences’ (1864, p. 19).

On idiomaticity

‘Many adults live abroad for years without ever attaining this power of expressing themselves idiomatically; and many teachers are staggered by their most advanced pupils’ total incapacity in this respect. The failure arises solely from their not having committed idiomatic sentences to memory at first.’ (1868, p.44)

On memorization:

‘To reproduce sentences verbatim, is to speak idiomatically; and therefore the genuine colloquial knowledge of a language is attained by repeated efforts of the memory, not by vigorous exertions of the reasoning faculties’. (1864, p.48)

‘When a man has committed to memory a few well selected sentences, each containing different constructions, and has acquired the power of putting them together in all their variations, one rapid perusal of the grammar will suffice to convince him that he is already in possession of the whole syntax of the language’ (1864, pp. 209-210).

‘In learning anything by heart, repetitions are indispensable, and the more they are distributed throughout the day, the smaller will be the number required to impress the foreign phrases on the memory’ (1870, pp 6-7).

On task repetition:

‘It is useful [for the learner] to frequent public places as a listener; to ask several people in succession for the news of the day after having carefully read it all beforehand; […] but especially to engage strangers in conversation in subjects which he has previously discussed with others, in order that he may repeat his own questions and observations, with additions and improvements. These second-hand conversations are by far the most instructive. (1864, p. 93)

On partial competence:

‘A language learned in miniature … may seem, at first sight, to be miserably defective; but a vast reduction of labour is effected by this plan, and it creates a great facility for the beginner in supplementing all his deficiencies’ (1864, p. 131).


Howatt, A.P.R. (with H.G. Widdowson) (2004) A history of English language teaching (2nd edition). Oxford: Oxford University Press.

Palmer, H.E. (1921) The Principles of Language-study. London: Harrap.

Prendergast, T. (1864) The Mastery of Languages, or the Art of speaking Foreign Tongues idiomatically. London: R. Bentley.

Prendergast, T. (1868) Handbook to the Mastery Series, New York: Appleton and Co.

Prendergast, T. (1870) The Mastery Series: French (new edition). New York: Appleton & Co.




30 responses

6 08 2017

Hi Scott,

While doing some background reading for a chapter on alternative current ELT methodologies (wink), I came across this article:

Lorch, M.P. (2016) A Late 19th-Century British Perspective on Modern Foreign Language Learning, Teaching, and Reform: The legacy of Prendergast’s “Mastery System” Historiographia Linguistica Vol. 43:1/2 pp. 175–208

Lorch develops the general argument, which you make, that the history of ELT methodology is a lot more nuanced than is generally supposed, and she uses Prendergast’s work to show how many current disputes were also discussed 150 years ago. Here’s a quote

“The most sagacious and systematic attempt to profit by the experience of childhood, though his interest was almost exclusively in learning to speak, was that of Prendergast…. The success of children in picking up languages was due, he held, to their following the light of nature. We have ignored that beacon and taken to grammar, whereas the child advances progressively by means of simple words in model sentences. Such words and sentences it was the object of Mr. Prendergast’s book to provide”.

6 08 2017
Scott Thornbury

Thanks, Geoff. That Lorch article sounds like one I should get hold of.

25 09 2017
Nick Prendergast

Fascinating post. Just out of interest, where did you source the photo?

25 09 2017
Scott Thornbury

It’s attributed to the Universityof California. Are you a relation?

26 09 2017
Nick Prendergast

Distantly, yes. His father Jeffrey was my great, great, great grandfather Guy Lenox Prendergast’s brother. If he were alive, we’d be cousins, I suppose. I know nothing about him other that what I’ve read.

26 09 2017
Nick Prendergast

Coincidentally, there are some similarities between us in that I too have lived out in Asia and also speak two or three Asian languages. In an ideal world, I would have liked to pursue a career in languages. Circumstances so far have dictated otherwise.

6 08 2017

Sorry, should have said the quote is from Ogden, 1930, cited in Lorch.

6 08 2017

I love this quote: ‘Usage is the only law. Usage constitutes the whole code’
Thank you for this post Scott!
Un saludo a todos 🙂

6 08 2017
Scott Thornbury

Thanks, Carlos. Yes, the choice of the term ‘usage’ seems incredibly prescient, as does the distinction Prendergast made between studying (i.e. learning) and acquisition.

6 08 2017
Heidi A. Karow

These insights are highly relevant to my current work, teaching basic English to those with little or no formal education.

I have sometimes thought of grammar as an attempt to make language logical like mathematics… but only that, an attempt. Exceptions, exceptions, exceptions…. always too many.

15 08 2017
Scott Thornbury

Indeed, Heidi, and, in fact, one branch of linguistics, as promoted by Chomsky’s mentor, Zellig Harris, sought ‘to apply the tools of mathematics to the data of language and establish the foundations of a science of language’ (to quote from Wikipedia). Chomsky himself said that ‘mathematical study of formal properties of grammars is, very likely, an area of linguistics of great potential’ (Aspects of the theory of syntax , 1965, p. 62). Great potential for linguists, perhaps, but not for learners or teachers!

6 08 2017
Peter Cox


6 08 2017

Thank you, Scott. Do you think this ties in with how my current ESP students (adults in Germany who need to use English with international tourists) seem happiest with what I call the “phrase book approach”? The members of each group want to learn a number of set phrases including very specific vocabulary and hope to vary them if they need to. This pragmatic approach is common to my nurses, shopkeepers, winegrowers, hospitality staff etc. regardless of whether their English is at A1 or C1 Level.

6 08 2017
Scott Thornbury

Interesting, Di. Sounds like some version of the lexical approach would cater for their needs/expectations. (See, for example, the post L is for (Michael) Lewis).

6 08 2017
Elka Todeva

Fascinating reading, Scott. Gracias for directing us to such true gems. If we spent more time studying the history of our respective fields, perhaps we wouldn’t be missing some great insights and our taking things to the next level would go faster and “deeper”. Thora Goldschmidt talked about personalized language in the 1920s; P. Strevens promoted the idea of porous classrooms almost 60 years ago; Alexander Balan outlined in the 1930s what later came to be known as Charles Fillmore’s Case Grammar; Roman Jakobson has offered us a goldmine still sitting for the most part unexplored. David Burkus, who has written a lot on the myths of creativity, is right perhaps in saying that nothing is truly new under the sun. The novelty comes from the combination or application, not the idea itself. (though to think out of the box, one needs to first get to know the box).The example he often uses – the printing press used the technology of the wine press. Enjoy your rose in Sant Pol while writing your next posting !

7 08 2017
Scott Thornbury

Thanks, Elka. Your comment serves to remind us that so much cutting-edge linguistic theory was kept hidden from the West because it was tainted with Marxist associations – I’m thinking particularly Vygotsky, but also Bakhtin. In fact, that second-to-last quote of Prendergast’s (about ‘second-hand language’) seems not a huge remove from Bakhtin’s view that all language use is bricolage – cuttings-and-pastings, he might have called it now.


7 08 2017

The chunking principle seems to be part of the orthodoxy now, and rightly so in my opinion, embedded in the CEFR specification (“formulaic language”), and ACTFL according to Nick Ellis in “Constructions, Chunking and Connectionism”.

Your post prompted me to re-read the section on Prendergast in Howatt (and Widdowson) 2nd edition. This quote really jumped out at me “To Prendergast, the crucial feature of language was the capacity of human beings to conceive the generative principle – it is an ancient principle in the philosophy of language – but few had seen it in psycholinguistic terms before, and certainly no-one had tried to apply it to language teaching materials”.

(Not be confused with the more recent “Generative school” who have no interest in real speech “performance” of course.)

The section on Claude Marcel has really got me going too!

15 08 2017
Scott Thornbury

Thanks, Patrick, for that quote . and your caution about over-applying the term ‘generative’. And, yes, Marcel was another 19th century ‘visionary’. Maybe I’ll blog about him later.

7 08 2017
Luiz Otávio Barros

This is eerily fascinating, Scot. I read this post twice what stuck out the most (to me, of course) was the extent to which, way back then, the man was able to see how blurred the lines between phrase learning, lexical priming and proceduralization / “accurate fluency” are. And yet, maybe it’s only me, but I can’t help but wonder why phrase learning (aka providing students with “islands of reliability” – to use your term – for short-term fluency) hasn’t yet fully crossed over into mainstream ELT. Maybe we’re afraid to help students commit ready-made language to memory because so much of what we do in ELT these days is meant to foster “creativity” and free expression? And there seems to be a tension, and maybe a trade-off, between saying things that have never been said before and using “second-hand” language to get stuff done? Maybe we still associate phrase learning with the kind of mindless drilling that ran rampant in the English900 days? Sometimes I wonder if the closest we’ve ever come to putting phrase learning at the core of our syllabuses/processes was in the late 70s, early 80s, when the Strategies series came along. A lot of what they used to call “functions” way back then we can “lexical phrases” now, except that now chunking is largely at the periphery of our syllabuses.

8 08 2017
Scott Thornbury

Thanks, Luiz. Yes, methodology has had a hard time dealing with the fact that language use is driven as much by the idiom (i.e. memorized chunk) principle as it is by the open choice principle – to use Sinclair’s distinction. Methodologies have either misidentified the chunks (as in Prendergast’s case) and/or overdone the memorization aspect (e.g. Audiolingualism). It seems to me that even the Lexical Approach had a hard time finding the balance – hence my much-quoted, but somewhat glib, critique that it was “all chunks and no pineapple”.

8 08 2017
Justin Willoughby

I’m not sure if this is what Thomas Prendergast (cool name!) was talking about, but when I first arrived in Spain with almost zero Spanish, I heard someone say ‘Hablando del Rey de Roma!ˋ(speak of the devil) when someone who eveyone had been talking about suddenly showed up. I didn’t consider the grammar of the phrase at all. I just memorized it and waited desperately for my chance to use it at the opportune time.

15 08 2017
Scott Thornbury

Thanks, Justin. I suspect most language learners who are exposed to ‘real-life’ input have experienced something similar. I guess the challenge in the classroom is in seeking to maximise the chances that this ‘holophrase’ learning occurs, while at the same time using these holophrases as a resource for abstracting the grammar.

11 08 2017

Hi Scott, Thank you for this interesting article which largely agrees with much of the ideas I am carrying about teaching English to my learners (level 3 adults/refugees UK). in my classroom I focus on blended teaching of grammar and lexical chunks or meaningful strings of words that are hooked to the memory i.e.oh that was a heavy rain not strong rain, this way learners advanced better in their speaking skill. However, their grammar competence is standing still,learners are fascinated with this methodology of chunking approach ( as Swan named it) and the idea that this is the shortest way to advance towards ‘nativelike fluency’ and I realized the attractiveness of this approach dominates other aspects of language and hold it sidelined!! I am afraid this will undermine language competency in some situations where more opened – ended aspects of language use are needed.
My question is: did you, or any other reader, have an idea of how can I solve this imbalance of language learning? how can I change my learners attitude?

15 08 2017
Scott Thornbury

Thanks, Abdullah (an apologies for delay in responding). I think that if your learners are predisposed to a ‘chunking approach’ (i.e. internalizing holophrases – see my comment to Justin above) at the expense of learning the grammar that ‘glues’ these holophrases together, then this may not be an entirely bad thing – especially at lower levels where – as Rod Ellis argues – overt teaching of grammar may be a waste of time, given that grammar only really kicks in once the learner has assembled an extensive lexicon (or ‘phrasicon’). At higher levels, getting learners to write connected texts is probably the single most useful technique for encouraging a focus on grammar, especially syntax.

15 08 2017
Jamie C


Great post.

Maybe it’s just me seeking out things that confirm my hunches, but there seems to be a lot online these days about chunks: their role, their choosing (taken from texts, decided before hand by the teacher, emerged, taken from corpara) and their value. I even had two students out of nine ask for more ‘useful phrases’ in a mid course NA this week. Naturally, two others asked for more ‘grammar points’ as well.

The big thing for me, if we take it as granted that learning chunks/adaptable phrases is helpful for students, is how to select them. Really not sure how to go about that in any kind of informed way.

15 08 2017
Scott Thornbury

Thanks, Jamie – the question of selection (of chunks for teaching) is still unresolved. In the 1990s Jane and Dave Willis adopted frequency as a guideline, particularly those chunks or structures that are associated with the most frequent words in the language, As D. Willis (2003) wrote ‘Many phrases are generated from patterns featuring the most frequent words of the language’ – indeed, the frequency of the words may owe a lot to the frequency of the phrases that embed them. The Willis’s often use the example of ‘way’ – think of all those common phrases in which it is the nucleus: on the way, in the way, by the way, I (don’t) like the way… etc.

19 08 2017
Jamie C

The thing is, let’s take frequency as the principle for selecting, and choose phrases containing the most frequent individual words as the basis of language to be taught. It’s still hard to know how to teach this language: with prepared materials incorporating as much of them as possible and then doing controlled practice activities to get students “using” them, or as a teacher just try to notice when the chances to feed these phrases in at the right moment. I tend to start reeling at the sheer volume of language that needs to be learned. And then there are still issues about topic vocabulary that may not be that common but that students actually need to talk about themselves and their interests.

Some problematic questions that I came across yesterday on G. Jordan’s blog related to this:

“How should the repeated exposure to massive numbers of lexical chunks be organised? Is frequency of occurrence in the biggest corpora the only criterion for the selection and presentation of lexical items, or are there others?

How do teachers make classroom sessions dedicated to “repeated exposure to the most frequent words in English” interesting and motivating?

How much input do learners need? How do Walkley and Dellar respond to the research findings which suggest that it’s unreasonable to expect FL classroom learners to remember even a small subset of what native speakers know?

Sinclair (2004: 282) warns of “the risk of a combinatorial explosion, leading to an unmanageable number of lexical items” and Harwood (2002: 142) warns against “learner overload”, insisting that “implementing a lexical approach requires a delicate balancing act” between exploiting the richness of fine-grained corpus-derived descriptions and keeping the learning load at a manageable level.

How do teachers help learners notice and store the thousands of lexical chunks which are required for a minimum level of proficiency?

Put another way, how do teachers help learners turn massive loads of input into an ability to use the language for effective communication?”

Obviously, not demanding answers to these questions but may be of interest to your readers checking out this post.


19 08 2017
Scott Thornbury

Thanks, Jamie. As I said in my response to Luis above, course designers, materials writers and teacher have had a difficult time trying to reconcile the different facets of proficiency, specifically the tension between item-leaning and rule-learning (and it may be the case that individual learners are differently disposed towards one or the other). Widdowson, as far back as 1990 (and three years before Lewis write The Lexical Approach), was already challenging the bias towards grammar teaching:

‘What is crucial for learners to know is how grammar functions in alliance with words and contexts for the achievement of meaning. The teaching of grammar, as traditionally practised, does not promote such an alliance. On the contrary, it is the formal properties of the device which are commonly given prominence. Words come in only as convenient for purposes of illustration… I would suggest that a more natural and more effective approach would be to reverse this traditional pedagogic dependency, begin with lexical items and show how they need to be grammatically modified to be communicatively effective.’

And he adds ‘within the category of lexical items, I include … formulaic patterns…. If they do figure so prominently in competence, it does not seem reasonable just to disregard their existence and leave their learning to chance… This would not imply that all of the many thousands of formulaic expressions would be expressly taught. The object would be to use them to develop learning procedures which would provide the basis for learners subsequently to process language and acquire packaged units for themselves.’

Widdowson, H. G. 1990. Aspects of language teaching. Oxford University Press, pp.95-96

19 08 2017

Jamie, I can’t comment specifically on the Walkley & Dellar/Sinclair/Harwood debate as I haven’t read any of it, but I think, in general, academics and teachers have different perspectives.

But the question about how to organise “the repeated exposure to massive numbers of lexical chunks” is at variance with Prendergast’s approach as quoted in Scott’s post, ie for Prendergast only a few compound chunks are required:

“When a man [or woman, or child] has committed to memory a few well selected sentences, each containing different constructions, and has acquired the power of putting them together in all their variations, one rapid perusal of the grammar will suffice to convince him that he is already in possession of the whole syntax of the language’ (1864, pp. 209-210).”

In practice, the chunking/formulaic/lexical approach has been embedded in many course books for a long time, with Jane & Dave Willis’s Collins Cobuild English Course (1988) probably one of the first in “modern” times.

I managed to track down a second hand copy of the Willises’ book recently. Formulaic chunks are just part of their approach, which also includes functions, notions and grammar.

Also, i know it’s obvious, so forgive me for saying it, but different people often need different chunks! (Even if the encapsulated grammar is the same).

Having said that, a book of graded chunks would definitely help me, I wonder if anyone knows of one?

8 09 2020
Evan Millner

If you want to try Prendergast out, I have been producing his Latin course as an audio course, closely following his instructions for presenting the material. A few of the lessons are available as sample lessons. https://www.latinum.org.uk/beginner/prendergast

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: