Classical Language Courses - DR Langslow

Notes on Some Aims and Assumptions of Classical Language Courses at University

D.R. Langslow

  1. Language Aptitude Testing
  2. The relation between active and passive competence
  3. The training of graduate students to teach undergraduate language-courses
  4. Summary

Under this rather pretentious title, I offer some brief thoughts on three related topics bearing on the teaching of Latin and Greek at university, which have occupied a number of us at Oxford over the last few years.[1] In the first two sections, I have tried to make my remarks as far as possible generally applicable to any classics degree-course with a language-based element. There is some straightforward description of current Oxford practice but, this apart, these notes should be read as a set of purely personal reflections, the reflections, moreover, of one who is in no sense an expert in these matters, who has indeed less experience in them than most of his audience/readership, and who hopes for feedback, both positive and negative, on any of what follows.

1. Language Aptitude Testing

An attempt to measure the ability to learn one of the classical languages quickly and accurately has since 1994 played a part in Oxford admissions-tests for single-language part-one classics courses ('Lit. Hum. Course II").[2]

In the public domain, language aptitude testing is about 70 years old and has been a major industry for more than forty years, its chief investors and consumers being the armed services and government departments, especially in the United States.[3] The LAT industry is driven by a desire to avoid more the spending of money on, than the causing of unhappiness to unsuccessful students; it rests on the long-held belief that 'facility in learning to speak and understand a foreign language is a fairly specialized talent (or group of talents), ... relatively independent of those traits ordinarily included under intelligence"' (Spolsky [n. 3], p. 128). In the inter-war years, intelligence tests had been relatively unsuccessful in screening for language-courses. Teachers in schools and universities reported significant numbers of failures in language-courses among the most intelligent students. On the other hand, special prognostic tests tried in the 1920s and 1930s, which generally correlated quite highly with intelligence tests, were often reasonable predictors of learning to read and translate. Their reported shortcoming, of being less good at predicting success in learning to speak a language in an intensive course, is probably not of serious concern to teachers of Latin and Greek. These early prognostic tests generally took the form of mini-lessons in an unfamiliar language followed by questions in or about it.[4]

In the view of one of the gurus in the field, John Carroll, expressed in 1960,[5] language aptitude consists of four distinct and measureable abilities:

  1. phonetic coding - the ability to code and remember an auditory phonetic signal
  2. grammar handling - the ability to recognize functions of words in sentences
  3. rote memorization - the ability to recall foreign-language items
  4. inductive language learning ability
(of which the Oxford test aims to test (b) and (d)). It is not immediately clear how one can cheaply and easily test (a), which is, in any case, probably irrelevant to an ability to learn to read and write Latin and/or Greek,[6] although it seemed important to include the (for English speakers) exotic phonological feature of distinctive vowel-length. The testing of (c) requires careful invigilation of candidates but can in principle be incorporated.[7]

Of course, success in learning a new language will depend on more than aptitude. We can all think of other factors which will crucially help or hinder the enterprise. There are, I think, no surprises in the set of factors listed and discussed by Carroll in a slightly later article,[8] three attributes of the student and two aspects of the teaching-situation:

  1. individual variables:
    1. verbal intelligence
    2. aptitude (time needed to learn)
    3. motivation (time and concentration applied to learning)
  2. instructional variables:
    1. adequacy of presentation
    2. time allowed for learning.
There is the further point - which amounts almost to another 'individual variable' - that aptitude generally increases with experience in language learning. This may be inferred from, inter alia, the fact that success in learning a first foreign language appears to be a good predictor of success in further languages, provided that the languages are of the same linguistic type. In our case, this means, of course, of inflectional type and with Indo-European grammatical categories.[9]

At this point it is quite reasonable to object: if a number of factors are relevant to success in language-learning, some of them uncontrollable and unmeasurable, and if verbal intelligence and success in a first foreign language may predict future success no less well than an LAT, then is the LAT game really worth the candle? This is a serious objection if resources for administering tests are limited, and it is probably true that, if, for whatever reason, a separate LAT cannot be used, the careful use of a combination of other indicators will be a good, perhaps equally-good, way of admitting to, or streaming within, language-courses. I would, however, make two points, neither of them original, one as a qualification of one of these 'other indicators' and the second in favour of a separate LAT.

First: the definition of past 'success' in a foreign language is obviously crucial. Few here, I believe, would see much value as an indicator of future success in intensive language-learning in even a starred 'A' grade in GCSE French, although this may be all that we have to go on by way of evidence of linguistic attainment. Crazier but true is the fact that even an 'A' grade in 'A'-Level Latin is, in itself,[10] of very limited value as evidence of successful learning of the Latin language and hence as a predictor of future successful engagement with, say, Greek.

Second, a positive point in favour of a separate LAT: an aptitude test can to some extent counteract differences in attainment occasioned by heterogeneous backgrounds, it can, so to speak, level the playing-field, or make it, if not level, at least sloping rather than terraced. It may even reveal unrealized potential within prima facie homogeneous sets, if, say, teaching-/ learning-styles in the first foreign language are uncongenial and cause under-performance in certain individuals.

In this area, in which the inexactitude of the 'science' is heightened by a lack of time and other resources for testing and assessment, Peter Green's conclusion in the excellent preamble to his University of York Language Aptitude Test[11] is both sensible and practicable:

"The sensible approach to pupil placement in foreign language courses would seem to be to base it on as much information as possible - previous attainment in a foreign language, general academic ability, available IQ information, language aptitude and not least the pupils' wishes".

In the 1950s, the United States Air Force used a two-stage LAT programme: a four-hour LAT (the Carroll-Sapon test: see n. 5) was used to screen applicants for a trial language-course (three days intensive), which was used in turn to screen for the full eight-month course. Notwithstanding the difference in scale between the USAF in the Cold War and a British university classics department today, this procedure may be applicable to those classics degree-courses in which language-based components are optional, for the purposes of attracting students to, advising against and streaming within language-courses.

2. The relation between active and passive competence

In both teaching and testing, the balance between active and passive use of a language will depend to a large extent on the purpose for which we are teaching the language. The key question here is of course what sort of linguistic competence are our students to have when they finish our course: (a) a smattering, (b) a reading knowledge or (c) the competence to read and write as far as possible like an educated native speaker? My guess is that most of us would go for (b), some for (a) and none for (c). I regard 'a smattering' and 'a reading knowledge' as rather subjective terms but as referring to different (sets of) points on the same scale.[12] The essential difference between these two and (c) is the presence/absence of active competence as an aim of the course. An insistence on (in some cases the possibility of) teaching and testing in active use of the language(s) is something that we classicists have by and large given up. Spolsky ([n. 3], p. 139) reports a similar development in modern-languages departments in US universities during the 1930s and notes:

When war began, the needs of the armed forces for soldiers fluent in the spoken languages of enemies and allies had showed up the major gap left by the decision of American schools and colleges in general to go along with the Modern [Foreign] Language Study's proposal[13] ... that the main objective should be the ability to read the language ... This literary goal was reinforced by the fact that language teachers in US colleges and universities were on the whole trained in, and carried out their own research on, the literature of the language they taught.

The question of the relation between the nature of language teaching/testing in a university and the research interests of its language-teachers, or rather decision-makers, is an interesting one for us as well. For whatever reasons, we, too, by and large are satisfied 'on the language side' if our students demonstrate some level of proficiency in reading. Don't worry: I am not about to ask, 'What will we do if there's another war?" This is just a roundabout way of raising the question what we mean by 'a reading knowledge'. Those of us who would admit the phrase in describing the main aim of their language course would possibly give a variety of definitions of what 'a reading knowledge' actually is. Spolsky ([n. 3], p. 86) quotes some interesting definitions of reading knowledge offered just before the war by thirty modern-languages departments in American universities to a questionnaire, the results of which were published in 1939:[14]

The main answers were: 'ability to read or translate with understanding or give the accurate rendering of a relatively difficult text, or a reasonably correct translation of a typical text without the excessive use of a dictionary' (seven responses); 'ability to read and understand without using a dictionary a given passage ... of normal difficulty' (four responses); 'ability to get the sense of a moderately difficult passage ... to read a text of average difficulty at sight ... to get the main ideas of a paragraph with its essential connotations ... to read with understanding texts of both narrative and of content' (eight responses); 'ability to use language as a tool' (three responses).

I have highlighted the second answer because, although there is room for debate on the meaning of 'normal difficulty', I think it is a good, concise definition of a reading knowledge of natural languages (Latin or Greek, for example!), and because it is, in a sense, the answer that we give as a professional community insofaras we teach and test passive competence by means of unprepared translation. But only in a sense, for the highlighted phrase implies to me a natural activity, something one does whenever the occasion arises, while writing an unseen is commonly a slow, artificial, unnatural exercise, and even a reliable producer of good unseens may not have the confidence to follow up a reference to a classical text of which no translation is ready to hand (it may not even occur to hrm that this is an appropriate use of time and of hrs knowledge of Latin or Greek). This kind of reading knowledge is very tentative, context-dependent, very passive, and it is easy to see why: Latin and Greek are difficult languages, the standard prescribed texts are difficult examples of difficult languages, and, even if the hard work necessary for learning the grammar has been done at school, not enough time is available at university for practising reading.

This is all leading to the paradoxical (though again not new) suggestion that, although our aim is to teach passive competence and we have insufficient time to do so, we should spend some of that precious time teaching and testing some elements of active competence. I believe this because I believe two propositions to be true (although I have no better than anecdotal evidence for either):[15] (a) active use of a language automatically improves reading knowledge; (b) active use is the quickest way of instilling secure recognition skills. In other words, (I am suggesting), using Latin or Greek will not only not divert our students from their goal of some level of reading knowledge, it will get them there in fewer hours.

In practical terms, I have in mind not lessons devoted to translating sentences from English into Latin/Greek but rather a constant intermingling of calls on active and passive skills, a steady shower of little tests in the active use of the language(s) against the background of a standard reading course. I mean tests or problems such as the following:

  1. supply the correct form of each word in brackets:
    Laetos (DIES) plures quam (TRISTIS) uidimus.

  2. fill the gap so as to give the required meaning:
    Altam fossam fecimus ..............(so that nobody)............. possit transire.

  3. transform the sentence in the manner indicated:
    Ad urbem eo ut panem emam: make this refer to past time
    Fratrem occidit et in siluas fugit: replace the first verb with a participle

  4. rewrite this ungrammatical sentence so as to make it grammatical:
    *Pollicebat pecuniam reddere. (He promised to return the money.)

  5. arrange the following words in a meaningful order and translate:
    canebat est qui puella secuta hominem

For illustrative purposes, these few examples deal with isolated sentences - actually, I do believe that isolated sentences have their place, e.g. when a particular construction or point of grammar is being taught, practised or tested. But, given our goal of a reading knowledge, little problems of this sort are probably at their best when incorporated in 'doctored' passages for reading, whether in class or for homework. Such texts may also contain e.g. ten deliberate errors for the students to spot and correct, or a section in which subordinate clauses have been turned into main clauses, or a section in which the order of the words has been altered and needs to be reconstructed in the light of patterns learnt last week/at the beginning of the present piece, and so on. Or again, in that all-important reading-skill of being able to group the words in sense-units in the order in which they appear on the page, I have found that good instruction and useful practice is given by using passages without punctuation (and capital letters!).[16] Then there is the cloze technique, and variations on it, illustrated in (f) with a short passage from Caesar which is missing most of its verbs. The task to fill the n gaps can be carried out either from a supplied list of n or 2n words, or from the student's own imagination:

(f) Interim ad Labienum per Remos incredibili celeritate de uictoria Caesaris fama ............................, ut, cum ab hibernis Ciceronis milia passuum circiter LX ............................, eoque post horam nonam diei Caesar peruenisset, ante mediam noctem ad portas castrorum clamor ............................, quo clamore significatio uictoriae gratulatioque ab Remis Labieno ............................ . Hac fama ad Treueros perlata, Indutiomarus, qui postero die castra Labieni oppugnare ............................, noctu ............................ copiasque omnes ............................ . Caesar Fabium cum legione ............................ in sua hiberna, ipse cum tribus legionibus circum Samarobriuam trinis hibernis hiemare ............................ et, quod tanti motus Galliae ............................, totam hiemem ipse ad exercitum manere decreuit.

I am arguing, I suppose, for the development and use of various sorts of 'super-unseens', to the end of making our students alert and critical readers of a 'live' text and more than mere passive receivers of a text that is straightforwardly given, and of training them to a level higher than that required by the writing of a weekly unseen, so as to achieve in a fuller sense the stated aim of reading and understanding without using a dictionary a given passage of normal difficulty.

If, finally, the course or module contains the option or requirement of writing complete sentences (connected or not) in Latin/Greek, it may be that practice in retranslation, ideally from prescribed texts/authors, will be useful preparation and have the advantage over invented sentences of serving a double purpose: (i) obviously, of testing active use of the language: a straightforward English version may fall back into straightforward Latin (which naturally earns full marks); (ii) of relating the exercise more closely to the development of a reading knowledge and the appreciation of the tendencies shown and liberties taken by the classical authors (and also (iii) of holding students' keen attention when the 'fair copy' is worked through: they love to see the original versions). The retranslation of complete sentences can, of course, grow out of gap-filling problems such as those in (b) and (f) above. It seems reasonable that students who have little or no experience in using a foreign language are best encouraged by being taken gradually from supplying endings and single words, through transformations, to unaided translation of whole clauses.

3. The training of graduate students to teach undergraduate language-courses

In Nottingham, at the CUCD's suggestion, I gave also a brief description of the Oxford seminar-course[17] which offers training and practice to graduate students in the teaching of 1st-year language-courses.[18] I reproduce the relevant section from my handout, for information on the way things have run over the last four years and on two areas where further development is felt to be necessary.

The pattern to date:

(a) week 1one 90-min. session to meet colleagues, set the tone, banish nerves; talk by course-leader on the basic ingredients of a good language-class; discussion of general and local aims and assumptions.
(b) weeks 2-8three 90-min. sessions, one each in 'Latin Syntax', 'Greek Syntax', 'Latin/Greek Reading Skills' [the primary aim of the last being technique].
(c) sessions2 x 30-min. simulated classes given in week 2 by the course-leader and Language Instructors from previous years, and then in weeks 3-8 by new 'trainees', those wishing to be Language Instructors in the coming winter terms (usually 1st-year graduate students); followed by group discussion of the strengths and weaknesses of each simulated class.
(d) JulyLanguage Instructors are selected on the strength of their performance in these practice/training sessions, and engaged to teach language-consolidation classes in the coming October/March.

Desiderata identified:

  1. monitoring: at present solely through feedback from students, informally through College Tutors, formally on two end-of-term confidential questionnaires; proposed introduction of classroom observation by suitably-qualified senior members to give encouragement and constructive criticism to Language Instructors and reassurance of ongoing quality control to Sub-Faculty.
  2. more authentic classroom-simulation: at present 'trainees' and course-leader are the pretend-class and give nearly always correct answers;
  3. possible introduction of undergraduates to the training sessions (or, failing this, of 'actors' aiming to simulate imperfect, unresponsive audience).

4. Summary

I said at the beginning that the three topics would be related. They may have seemed quite disparate as things went along, so let me close by summarizing some of my main points and letting the linking thread emerge which I did not have time to spin in Nottingham.

  1. For many, but probably not all, branches of classical studies, a reading knowledge of the languages is important, and in many, but not all, classical degree-courses a language-course is either required or available.

  2. It is in the interests of all concerned to make some estimate in advance of the language-course of the likelihood of a student's speedy success in learning Latin or Greek from scratch.

  3. To this end a short written language aptitude test can probably furnish useful evidence to complement other indicators (such as verbal intelligence, motivation, earlier success with intensive language-learning).

  4. The language-course itself needs naturally to have a clear set of aims and objectives in terms of the linguistic competence that it can be reasonably expected to give in return for students' commitment to it.

  5. The all-too-little time available, especially at university, for learning and practising reading in Latin/Greek is, paradoxically, most efficiently used by developing a combination of passive and active linguistic skills, even if the stated aim of the course is some level of purely passive competence.

  6. If it is true that some active skills are necessary for effective language-teaching, at any level, then it is doubly important that they be taught in university courses, since increasingly university teachers of Latin and Greek will have had their first contact with (one or both of) the classical languages as undergraduates.

David R. Langslow

Wolfson College, Oxford


[1] I am grateful to the editor, Nick Lowe, for accepting what is little more than a hasty write-up of my Nottingham handout and marginal notes made for oral delivery. I should like to thank also the CUCD standing committee, especially Alison Sharrock, for the invitation to be part of the panel; John Richardson, for his hospitality on the day and for chairing the session so catalytically; my fellow-panellists, Barbara Bell and Anthony Bowen, and all who contributed to the informative and thought-provoking discussion.

[2] A recent 60-minute Oxford LAT is appended to these notes. In '94 and '95, the test was set in 30-, 60- and 90-minute versions; from '96 a single 60-minute paper will be set. It is, of course, much too early to assess the predictive value of this test. I am very grateful to Peter Green, formerly of the University of York (see the end of 1. below), for commenting on these tests: he is generally encouraging about section II but expresses doubts, which I share, about section III, which needs further development. Section I is just a friendly loosener.

[3] For a recent history see Bernard Spolsky, Measured Words (Oxford 1995), chapter 7.

[4] Quite independently, this is the pattern of section II of the Oxford LAT.

[5] J.B. Carroll, 'The prediction of success in intensive foreign language training (final revision)' (Graduate School of Education, Harvard University 1960): this is the report submitted at the end of about eight years of research on aptitude testing at the Laboratory for Research in Instruction, at the Graduate School of Education at Harvard University; see Spolsky [n. 3], p. 128. The work yielded, among other things, J.B. Carroll - S.M. Sapon, Modern Language Aptitude Test, New York 1959.

[6] This may be quite wrong - and I certainly do not mean to question the value as a teaching-aid of reading aloud in an ancient language.

[7] It appears, for example, in Paul Pimsleur, Pimsleur Language Aptitude Battery, New York 1966.

[8] J.B. Carroll, 'The prediction of success in intensive foreign language training', in R. Glaser (ed.), Training Research and Education, Pittsburgh, Pa 1962, 87-136.

[9] The invented languages in the Oxford LAT are, for all their barbarous appearance, very like Latin and Greek in their relevant features.

[10] Much will depend on the options taken, the course used at school, the extent to which the nitty-gritty of the language has been confronted or avoided.

[11] In Swedish, for use in the middle school years.

[12] I intend no disrespect at all by 'a smattering' - quite the reverse: as a comparative philologist, I know this level of linguistic competence at first hand and value it highly. A smattering is surely the right aim if time-constraints render unrealistic the attainment of a reading knowledge. I take it, though, that the essential feature of a good smattering is the enabling of the student to learn more of the language with confidence in private when the language-course is finished, and that this in turn depends on a nodding acquaintance with the whole grammatical system and on practised familiarity with the basic tools (dictionaries, grammar-books, collections of texts, etc.), especially those relevant to any special purpose (e.g. inscriptions, coins, mediaeval history) that may have prompted the student to take a short language-course with this sort of aim.

13 As the result of an enquiry conducted from 1924 to 1927. See Spolsky [n. 3], pp. 41-6, esp. 42: 'the members of the Committee of Modern Foreign Language Study appear in recent language testing histories as the villains who discarded the direct method rather than as the realists who saw no way to increase the amount of time students would give to language learning, and as the researchers who established the basis for empirical study of language teaching and testing.'

[14] A.L. Frantz, 'The reading knowledge test in the foreign languages: a survey', Modern Language Journal 23 (1939), 440-6.

[15] I should be very grateful for references to relevant published research in this area.

[16] I owe this idea to Nick Lowe.

[17] A most valuable and effective initiative (I hope I may say so), devised by David Raeburn and led by him each summer term from 1993 to 1996.

[18] These are courses in revision and consolidation for post-'A'-Level students, offered to Colleges and administered by the (Sub-)Faculty. The four courses, in Latin / Greek Syntax / Reading Skills, run each for an hour a week for the first two terms (= 16 weeks) and are taught mainly by graduate students, in streamed groups of about 10 for syntax and about 5 for reading skills; undergraduates sit a short written test in accidence and syntax at the start of their third term, i.e. after the consolidation course plus the Easter vacation for revision.

CUCD Bulletin 25 (1996)
© David R. Langslow 1996

Bulletin ContentsCUCD HomeRoyal Holloway Classics Dept