Grice's Descriptivism

It is one thing to speak Latin, and another to speak grammar. (Quintilian 1920-22, vi.27) 

Grammar seems to be permanently split into two camps: one of theoretical, philosophically oriented grammarians; the other of pedagogically engaged, and hence prescriptive, grammarians. A teaching grammar necessarily prescribes what linguistic structures are to be learned. 

It must demonstrate an approved model of grammatical usage to establish criteria for good usage and good style. In antiquity the standards for language excellence were the great poets like Homer, Virgil, and Horace; dramatists like Aeschylus, Euripides, Plautus, Seneca, and Terence; orators and rhetoricians such as Isocrates, Demosthenes, Cato the elder, Cicero, and Quintilian. The literary greats were normally canonized only when long dead: good style was invested in revered models from the past. Within the western world, from the fourth century to the nineteenth, this reverence for past authority in literary and grammatical excellence was transferred to the language of the Bible and the Church fathers. The celebrated originators of school grammars of Latin, such as Donatus and Priscian, were also treated with reverence in the middle ages; it was, however, acceptable to criticize their statements on grammar and the limitations of their grammars as it never was tolerable to criticize the language of the Church, which set a new standard for pure Latin. Our chapter begins with a review of pedagogical grammar in the Europe of the early middle ages, which established a tradition of prescriptive grammars that has lasted until today. 

During the renaissance, or a little before, interest shifted from sole concern with the grammar of Latin (because it was the language of the Church) to writing grammars of vernacular languages that were modelled on existing word-and-paradigm grammars of Latin. European interest in vernacular grammar coincided with, and was probably motivated by: * the rise of Protestantism and the consequent translation of the Bible from Latin into vernacular languages; * the growth of printing, which took book production out of monasteries and made books widely accessible; and * concomitantly, the emergence of a nationalistic merchant class. Nationalism is usually accompanied by exaggerated pride in the national language: the language which is believed to best express the unique national culture and ‘genius’ of the people. Hence, there is an interest in the purity of the vernacular language expressed through prescriptions of ‘correct’ grammar and usage which create doctrines that apply to a notional ‘standard language’. In 1697 Daniel Defoe was moved to propose a ‘Society’ on the model of the Académie Française: The Work of this Society shou’d be to encourage Polite Learning, to polish and refine the English Tongue, and advance the so much neglected Faculty of Correct Language, to establish Prescriptivism from the early middle ages on 129 Purity and Propriety of Stile, and to purge it of all the Irregular Additions that Ignorance and Affectation have introduc’d; and all those Innovations in Speech, if I may call them such, which some Dogmatic Writers have the Confidence to foster upon their Native Language, as if their Authority were sufficient to make their own Fancy legitimate. (Of Academies, Defoe 1697: 233) In this chapter we look at prescriptivism in eighteenth century Britain as an offshoot, if not a continuation, of the prescriptivism of school grammar in the early middle ages. Behind the prescription is the notion of a standard for English – more accurately, as many notions of a standard as there were grammarians, albeit with common threads. 

Alongside the pedagogical and prescriptive grammars, there was between the eleventh and the eighteenth centuries a new interest in the theory of grammar, which we examine in Chapter 8. It had some effect on the pedagogical tradition, as we shall see, but it focused on and philosophized about the essential nature of grammar. The eighteenth century doctrines of ‘correctness’ in English usage continued to affect the teaching of language throughout the nineteenth century and well into the twentieth. Indeed, there are very similar notions of ‘correctness’ around today. Figure 7.1. Europe in the early middle ages. 130 The Western Classical Tradition in Linguistics The early middle ages They were called the ‘middle ages’ by renaissance scholars who viewed the period between the flowering of the classical world and the rebirth of learning in the fifteenth century as an era of darkness and philistinism. It was a conservative epoch where rote learning remained the norm; but it was not without innovators. The fall of Rome to the Visigoth (German) invasion resulted in the loss of much classical literature. The philosophical schools of Athens were closed by the Byzantine Emperor Justinian I (ruled 527–65) in 529. The eastern Roman world, the repository for ancient Greek works, was difficult to access from the west and north; and by the late sixth century the Byzantine state was confronting many of the same threats that had destroyed the western empire in the fifth century. Barbarians pressed upon it from beyond the Balkan frontier while others manned the armies defending it. Wealth accumulated during the fifth century had been expended; and there were few native Romans living there. The Byzantine empire avoided the fate of Rome by requiring its barbarian subjects to accept Orthodox Christianity and the emperor’s authority. Christianity was a fragile veneer that often cracked in moments of crisis, and loyalty to the emperor was often forsworn. Nevertheless, the Christian faith and the ecclesiastical institutions defined in the sixth century proved more cohesive and morale boosting than the pagan culture of the western empire. Until the late eighth century, today’s Romance languages were more regional dialects than distinct languages. 

The situation was probably similar amongst Germanic peoples; missionaries from Anglo-Saxon England could readily communicate with Saxons and Austrasians in the east of the Frankish Empire. As time passed, the Latin of antiquity ceased to be anyone’s mother tongue. Where Donatus could assume that his readers were familiar with Latin, the grammarians and commentators of the early middle ages had to help those for whom it was a second or foreign language. Instead of studying grammar to improve understanding and delivery of the classical Roman texts, it was a case of studying Latin grammar in order to understand Christian Scripture. The earliest Latin biblical texts were in vetus Latina “old Latin” and not only gave many variant readings for the same source text but abounded in solecisms and calques on Greek and Hebrew from the Septuagint.1 The new translation in St Jerome’s Vulgate (382–405) quickly became the definitive biblical text (we can ignore the fact that there was more than one version). 

Throughout western and northern Europe a knowledge of Latin was needed to understand Scripture and exegetical works, Church liturgy and ritual, or to undertake ecclesiastical duties such as delivering a sermon. Latin was the lingua franca of Europe. The grammatical works of Donatus and to a lesser extent Priscian were studied, copied, commented on, and elaborated. The principal innovation was the Christianizing of grammars. Classical Latin was written by pagans, and many Churchmen believed that it could not be studied in detail without barbarizing the 1. Supposedly the Hebrew Bible was translated into Greek independently by 72 scholars in 72 days; and these holy men were so inspired by God that their translations did not differ from one another by a single word. In fact, the Septuagint was compiled at various periods between the third and first centuries BCE. By Jewish and patristic tradition there were (?coincidentally) 72 languages in the world. Prescriptivism from the early middle ages on 131 student. 

In any case, the Latin of the Church fathers was more highly valued because it had God’s imprimatur. Pope Gregory the Great (590–610) in his Epistola missoria ad Leandrum Hispalensem declared, ‘I deem it most unworthy to restrict the words of the heavenly oracle under the rules of Donatus.’ The ninth century French Abbot Smaragdus of St Mihiel-surMeuse (fl. 809–19) based his grammatical commentary on Donatus, but held the authority of the Vulgate to be superior. In all these things we do not follow Donatus, because we hold the authority of Divine Scripture to be greater. Indeed we do not deny that cortex, silex, stirps and dies have common gender; however we are taught by the authority of the Scriptures that radix, finex and pinus are of feminine gender. (Smaragdus 1986 4T, 121–25)2 Names and notions were chosen from Scripture and the Church milieu to replace those of pagan Rome. The pedagogical method of Donatus and of Priscian’s De Nomine and Partitiones (Priscian 1961a; b) was hijacked and Christianized as in the early eleventh century Beatus Quid Est? “What is holiness?” (from England) with its uplifting vocabulary. For instance, the lists of comparatives: castus [“chaste”], castior [“more chaste”], castissimus [“most chaste”]; probus [“upright, worthy”], probior, probissimus; iustus [“just”], iustior, iustissimus; sanctus [“sacred, saintly”], sanctior, sanctissimus; doctus [“learned”], doctior, doctissimus; beatus [“blessed, happy, holy”], beatior, beatissimus (Bayless 1993: 90). It was generally accepted that the standard for grammars should be founded on three criteria: * auctoritas – the authority of an accepted model (Virgil, Horace, Plautus, Cicero, Varro, Donatus, Charisius, Priscian); auctoritas came to be confused with vetustas “antiquity”; * ratio – rational explanation in a systematic account consisting of definitions and rules (regulae, analogia); * consuetudo or usus – contemporary usage. Of these criteria St Augustine (354–430) judged ratio the least important (contrary to Apollonius) and consuetudo the most important (Augustine 1861). 

But St Isidore of Seville (560–636) reveals a lack of confidence in consuetudo, an awareness that Vulgar Latin was not like the classical Latin of the Roman republic or early empire, but had become ‘corrupted’ by solecisms and barbarisms (Etymologies IX.i.6–8, Isidore 1617; Brehaut 1912). That was one good reason for the heavy reliance on old faithfuls like Donatus and Priscian (Priscian was accepted as a native speaker of Latin despite his North African origin and relocation in Byzantium). The works of these masters were reorganized and Christianized and commented upon, but there was no significant advance in grammatical theorizing or descriptive practices during the early middle ages. There was, however, development of word-and-paradigm introductory school grammars (primers) that virtually ignored syntax and established a pedagogical tradition that persisted until the twentieth century. 2. See also 4T 146, 152; 8T 115–20; 9T 52–57; and 5T, 62–65, where Smaragdus favours the use of a singular for scala “ladder, staircase” in Genesis against Donatus’ prescribed plural. 132 The Western Classical Tradition in Linguistics Isidore of Seville, encyclopaedist Greek science had fought against superstition and engaged in empirical observation and rational interpretation of reality. In the early middle ages attention centred on the spiritual world of neo-Platonist Christianity.3 For example, Isidore wrote history and many theological works, including one on the canon law of the Spanish church (Isidore 1862b); he wrote treatises on linguistics (Differentiarum Libri “Books of differences”, Isidore 1862a), on natural science and cosmology (De Ordine Creaturarum “On the order of creatures” and De Natura Rerum “On the nature of things”), and on the mysticism of numbers (Liber Numerorum, Isidore 1862c). Outstanding among Isidore’s literary production were the 20 books of his Etymologiae (hereafter Etym.; Isidore 1850), which compiled extracts from the works of previous encyclopaedists, specialists, and various Latin writers as well as his own observations. 

Almost 1,000 manuscripts of the Etymologies are still in existence. Isidore fuses education in the seven liberal arts with more general encyclopaedic data, using ideas from Aristotle (though he had no Greek and had to rely on translations), Nicomachus of Gerasa (numerologist, fl. 100), Porphyry, Varro, Cicero, Suetonius, Moses, St Paul, Origen (theologian, 185–254), St Jerome, St Augustine, and Donatus. Isidore plagiarizes Magnus Aurelius Cassiodorus (c. 490–585); especially De Artibus ac Disciplinis Liberalum Litterarum (Cassiodorus 1865). From the time of Varro, Latin scholars had preferred to summarize all knowledge rather than specialize. Varro and Pliny the Elder (23–79 CE) cribbed from original texts; but the encyclopaedists who followed them drew further and further away from primary sources. There was little attempt to match reported statements with observations of reality: what was significant was the authority of the source and not pragmatic validation of its assertions. If there was ever a difference of opinion, Isidore preferred the authority of the Bible or a Church father. He confidently asserts that Origen, Jerome, and Augustine surpassed all pagan authors in genius, wisdom, quality, and number of works (Etym. VI.vii.2–3). Rather curiously, he says: We have God’s apostles as authorities. […] 

And so even if an angel from heaven shall preach otherwise, let him be anathema.4 (Etym. VIII.iii.3). Isidore asserts that ‘Hebrew is the mother of all languages and alphabets’ (Etym. I.iii.4), sourcing this from St Jerome (Epistolae XXIX, XXX in Jerome 1864: 441ff). The diversity of languages arose after the flood, at the building of the tower [of Babel5 ]; for before that […] there was one tongue for all peoples,6 which is called Hebrew. (Etym. IX.i.1) 3. Neo-Platonism is an amalgamation of Plotinus’ (c. 205–70) interpretations of Plato, Porphyry’s (c. 234–305) edition of Plotinus’ Enneads, the Septuagint, Christian, Gnostic, and Jewish teachings. The Cosmos is the flawed expression (presentation) of the Divine (Platonic) Idea. See http://www.iep.utm.edu/n/neoplato.htm. 4. Translations of Isidore are taken, sometimes slightly adapted, from Brehaut 1912. 5. Genesis 11: 4–9. 6. Genesis 11: 1. Prescriptivism from the early middle ages on 133 Adam, of course, spoke Hebrew (Etym. XII.i.2). But it seems that God speaks to a person in their own language, whatever it may be (Etym. IX.i.11). There are three sacred languages, Hebrew, Greek, and Latin […] For it was in these three languages that the charge against the Lord was written above the cross by Pilate.7 (Etym. IX.i.3) The strength of Isidore’s beliefs in the authority of the Bible and its disciples was typical of scholars in the period between late antiquity and the renaissance – if not well into the nineteenth century. Like many of his contemporaries Isidore was a mystic. He perceived patterns and connections where we cannot. For instance, between the knees and the eyes: [M]an in his beginning and first formation is so folded up that the knees are above [against the eyes], and by these the eyes are shaped so that there are deep hollows. […] Thence it is that when men fall on their knees they at once begin to weep. For nature has so willed that they remember their mother’s womb where they sat in darkness, as it were, until they should come to the light. (Etym. XI.i.109) And there was the magical number 22: [I]n the beginning God made twenty-two works. […] 

And there are twenty-two generations from Adam to Jacob, from whose seed sprang all the people of Israel, and twenty-two books of the Old Testament as far as Esther, and twenty-two letters of the [Biblical Hebrew] alphabet out of which the doctrine of the divine law is composed. (Etym. XVI.xxvi.10, see also I.iii.4 and VI.i.3) The Jewish historian, Flavius Josephus, wrote (c. 90 CE) that Jews recognized 22 books (Josephus 1737); today Jews recognize 24 books. St Augustine followed the Septuagint and listed 44 books. There are 27 letters in Modern Hebrew. Isidore’s approach to etymology is also mystical: ‘His idea was that the road to knowledge was by way of words, and further, that they were to be elucidated by reference to their origin rather than to the things they stood for’ (Brehaut 1912: 33). In other words, Isidore made no distinction between natural and philological sciences; he was an out-and-out ‘naturalist’. A knowledge of etymology is often necessary in interpretation, for, when you see whence a name has come, you grasp its force more quickly. For every consideration of a thing is clearer when its etymology is known. Etymologies are given in accordance with cause, as reges from regere [“to rule”], that is recte agere [“do what is right”]; or origin, as homo [“man”] because he is from humus [“earth”]; or from contraries, as lutum [“mud”] from lavare [“to wash”] – since mud is not clean – and lucus [“sacred grove”], because being shady it has parum luceat [“little light”]. (Etym. I.xxix.2, 3; cf. I.xxxvii.24) 7. Cf. John 19: 19–20. The titulus crucis in Hebrew was ??????? ??? ?????? ??? YESHU HANOTZRI MELECH HAYEHUDIM (reordered from the right-to-left Hebrew script); the Greek was ?????? ? ????????? ? ???????? ??? ???????? and in Latin IESUS NAZARENUS REX IUDAEORUM (whence INRI). All of these would have been in scripta continua. The three languages are vouched for in Luke 23: 38, but the wording is different: ‘This is the King of the Jews’. Matthew 27: 37 and Mark 15: 27 don’t mention the three languages, and the wording is different in each, though all gospels report ‘King of the Jews’. 134 The Western Classical Tradition in Linguistics Ossa [“bones”] are named from ustus [“burned”], because they were burned by the ancients, or as others think, from os [“the mouth”], because there they are visible [teeth], for everywhere else they are covered and concealed by the skin and flesh. (Etym. XI.i.86) Apes [“bees”] are so called because they hold to one another by the feet [pes “foot”], or it may be because they are born without feet. […] In a correct sense apes are so called because they spring from boves [cattle] as hornets from horses, drones from mules, wasps from asses. (Etym. XII.viii.1, 2) There is, of course, the medieval belief that the Earth is at the centre of the universe. The earth is placed in the middle region of the universe, being situated like a centre at an equal interval from all parts of heaven. 

In the singular number [orbem], it means the whole circle, in the plural [terrae] the separate parts. And reason gives different names for it; for it is called terra for the upper part where it suffers teritur [“erosion”]; humus from the lower and humid part, as, for example, under the sea. Again, tellus, because tollimus [“we take”] its fruits. It is also called ops [“power”] because it brings opulence. It is likewise called arva [“ploughed”], from arando [“ploughing and cultivating”]. (Etym. XIV.i.1) On the basis of a twenty-first century understanding of the term etymology, most of Isidore’s etymologies are absurd. However, as I said in Chapter 4 when discussing Varro’s etymologies, the ‘naturalist’-inspired understanding of the term in antiquity, which remained in force throughout the middle ages and is still with us,8 would be better glossed in today’s terms as “lexical semantics”. To be more precise, the aims of the ancient and medieval etymologist were, mutatis mutandis, similar to those of today’s students of lexicology in trying to map lexical relations in terms of both morphological form and meaning. The assumptions and methods are very different, but the aim is very similar. The modern understanding of etymology is “tracing the diachronic development of a linguistic expression via the changes in its meaning and form, identifying related expressions within the same and other languages, and reconstructing the proto-item” – we might loosely gloss this as word-history. This is just one specialized aspect of lexicology, but for Plato, Varro, and Isidore (among many others) etymologia was the whole of lexicology and more. Here are some more of Isidore’s etymologies. In the first three he identifies correlations within a semantic field9 among words that appear to be formally related and therefore might plausibly be recalled as a group. Disciplina [“teaching”] takes its name from discendo [“learning”]: hence it can also be called scientia [“knowledge”]. So scire [“to know”] is so called from discere [“to learn”] because 8. Robert Southey (1774–1843) suggested, perhaps not light-heartedly, that lass derives from alas, sighed at the thought that the girl ‘would in time become a woman, – a woe to man!’ ‘[B]ecause by woman was woe brought into the world’ (Southey 1834-47, 7: 76f); but the OED entry 1.k for woman shows that many seriously derive the word from woe to man. 

The late twentieth century feminist neologism herstory, dissimilating it from history, is based on false etymology. It is said that male hysteria is an oxymoron because Greek hystera means “womb”. And so forth. As Kate Burridge has pointed out to me (p.c.), these false etymologies demonstrate a kind of prescriptivism. 9. The semantic field of a word is determined from the conceptual field in which its denotatum occurs; it is structured in such a way as to mirror the structure of the conceptual field. See Chapter 13. Prescriptivism from the early middle ages on 135 none of us knows unless he learns. Otherwise it is called disciplina because it is discitur plena [“fully learned”]. (Etym. I.i.1) They are called litterae [“letters”] almost like legiterae because they offer an iter [“path”] to the legentibus [“readers”] or because they are in legendo iterentur [“repeated in reading”]. (Etym. I.iii.3) The word princeps [“chief”] gets its meaning from capiendi [“taking”] because he primus capiat [“takes the first place”] just as municeps [“official”] from one who munia capiat [“undertakes public duties”]. (Etym. IX.iii.21) In the next three quotes Isidore is attempting to explain the meanings of oratio, articuli, and verbum also by identifying correlations within a semantic field. Oratio [“sentence, utterance”] is almost oris ratio [“reason of the mouth”]. (Etym. I.v.3] Articuli [“articles”] are so called because they artantur [“are pressed together with”] nouns. (Etym. I.viii.4) Verbum [“word, verb”] is so called because the verberato [“vibrating”] air resounds, or because this part of speech versetur [“is used”] frequently. (Etym. I.9.1) 

Etymologia included finessing the original forms and meanings of words in what would today be called the proto-language; it included word history, but it was a wider concern with relationships among the vocabulary items in the language with the focus being always on nouns or names (since ????? and nomen mean both “name” and “noun”). We would today include part of this enterprise under morphological analysis,10 but for the ancients it had a semantic aspect that makes a term like lexical relations more appropriate. The relationships identified are often mistaken, but the meaning of etymologia is, nonetheless, “lexicology” or “lexical relations” and occasionally “lexical semantics”, rather than what we today call etymology. Isidore is a prime example of scholarship in the early middle ages: existing knowledge was not extended, but compiled and rearranged. As Brehaut 1912: 32 says of Isidore’s endeavour: ‘Secular knowledge had suffered so much from attrition and decay that it could now be summarized in its entirety by one man.’ We are sometimes struck by the absurdity of his claims and his gullibility; here is just one example: The Blemmyes, born in Libya, are believed to be headless trunks, having mouth and eyes in the breast; others are born without necks, with eyes in their shoulders. (Etym. XI.iii.17) Perhaps the most perceptive remark that Isidore made on linguistics is the unoriginal ‘Grammar is the science of speaking correctly, and is the source and foundation of literature [liberalium litterarum]’ (Etym. I.v.1). 10. In each of Ann Fisher’s A New Grammar (Fisher 1753) and Lindley Murray’s English Grammar, discussed later, one of the four sections was ‘Etymology’, which ‘treats of the different sorts of words, their derivation, and the various modifications by which the sense of a primitive word is diversified’ (Murray 1795: 19). What they in fact did in this section was discuss the parts of speech and their various subcategories in a word-and-paradigm grammar that might today also be called morphology. 136 The Western Classical Tradition in Linguistics The insular grammarians Although Isidore thought Britain ‘beyond the circle of lands’ and ‘that the Britons were socalled according to the Latin because they are bruti [“stupid”]’ (Etym. IX.ii.102),11 the most influential grammarians of the early middle ages were of British or Irish origin: ‘It was on the Continent that almost all the Insular grammars were studied and copied’ (Law 1982: 98).12 

In Ireland, Latin scholarship combined with bardic tradition to produce the medieval bardic tracts such as the Auraicept na n-Éces “The scholar’s primer”, part of which dates to the seventh century and is the oldest study of a west European vernacular. Ireland was in the forefront of Christian civilization until the Scandinavian invasions of the ninth century. The scotti peregrini, wandering Irish scholars, were influential throughout Britain and the rest of Europe; some, such as Anonymus ad Cuimnanum,13 Sedulius Scottus and Johannes Scottus Eriugena were grammarians or commentators on grammar. Viking raids in the late eighth and ninth centuries in Northumbria, East Anglia, Kent, and Wessex led to many Danes settling in England; Norse kings reigned intermittently in York until 954. King Alfred, who kept the Norsemen in check, saw the Viking menace as God’s punishment; so he encouraged a return to Latin learning to strengthen the Church while at the same time championing literacy in Anglo-Saxon. Few in Anglo-Saxon Britain knew much Latin until after 669, when the learned monks Theodore and Hadrian arrived to re-establish Latin scholarship. Only Donatus’ Artes, Priscian’s Institutio de Nomine (hereafter, De Nomine), and the first book of Isidore’s Etym. were widely known in Britain before the carolingian renaissance of the late eighth century. Aldhelm and Virgilius Maro, both possibly British, knew Priscian’s Institutiones Grammaticae. St Aldhelm of Malmesbury (c. 639–709), who claimed to have been instructed by Hadrian, wrote on Latin metrics but not grammar. ‘Aldhelm’s works are notorious for their highly elaborate, syntactically complex and lexically recherché Latinity’ (Law 1997: 190); they were influential in eighth century England. Virgilius Maro (Law 1995b) was a seventh century parodist who invented novel ‘Latin’ words and created a mythical set of grammarians.14 He was an original thinker and much quoted as a serious grammarian, though today it is difficult to see him as that. Undoubtedly British were Tatwine, Boniface, and Alcuin. St Tatwine (d. 734 when Archbishop of Canterbury) wrote the Ars Tatuini (Tatwine 1968), probably before 700, when teaching at the monastery at Breedon-on-the-Hill, Leicestershire. 

The detailed elementary grammar (primer) is based on Donatus’ Artes. Donatus offers only a very incomplete account of Latin inflectional morphology, but it was sufficient for his native-Latin-speaking audience. He used one third conjugation verb legere 11. According to Hiob Ludolf in a letter to Leibniz c. 1700, Britannia and Britain derive from Bretanae “region of tin” (Waterman 1978: 27). 12. This chapter owes a lot to the scholarship of Vivien Law. 13. Anonymus ad Cuimnanum 1992 has Hibernian properties and is addressed to an Irishman. 14. Virgilius’ writing makes one think of James Joyce and J.R.R. Tolkein. He had a very patristic notion of derivation starting with masculine nouns, e.g. glorificatio “glorification” ? glorifico ? “glorify.1.SG”; gloria “glory” ? gloriosus–a–um “glorious” (Amsler 1989: 202). Prescriptivism from the early middle ages on 137 to demonstrate the terminology applicable to describing its various forms; gender, though not declension, is covered by the five nouns magister (“master”, M), musa (“muse”, F), scamnum (“bench”, N) sacerdos (“priest(ess)”, M or F) and felix (“happy”, M or F or N). No verbs of the other conjugations or nouns of the fourth or fifth declensions are described. The Constantinople-based Greek student of Priscian, Eutyches, wrote the Ars de Verbo (Eutyches 1961), an alphabetically organized form-based account of verbs which supplements Donatus. Priscian’s De Nomine identifies the noun declensions and was the basis for many revisions of Donatus during the early middle ages. Tatwine utilized Eutyches and Priscian while taking his vocabulary from ecclesiastical literature; his other sources include Martianus Capella (fifth century CE, Capella 1983), Isidore’s Etym. Book I, Pompeius (Pompeius 1961, a fifth to sixth century North African commentator on Donatus), and the Ars Asporii (Asporius 1961). The Ars Asporii, probably composed in Gaul before 650, is a Christianized Donatus-based primer, with more numerous paradigms and more copious examples than Donatus, but which shows no knowledge of Priscian’s De Nomine, and does not identify declensions. The Ars Tatuini was unknown to the west-countryman St Boniface (Uynfreth Bonifatius, 675–754) when he wrote the Ars Bonifacii (Boniface 1980) at Nhutscelle (Nursling, Hampshire) before he left on a mission to Germany in 716. Concerned with Latin accidence, Boniface has one paradigm for each of the five declensions. He relies heavily on (indeed, plagiarizes chunks of) the third to fourth century Greek commentator Phocas’ form-based, alphabetically arranged Ars de Nomine et Uerbo (Phocas 1961). The Ars Bonifacii is a good primer that pays more attention to anomalies than other elementary grammars. Boniface’s Praefatio ad Sigibertum “Preface to Sigeberht” gives a revealing account of his intentions, assumptions, and procedures. I skimmed the works of a number of grammarians in cursory examination and carefully assembled the rules that are the most useful and essential in composition for each of the eight parts of speech. Where the meaning I lit upon was somewhat shadowy, in matters dealt with cursorily by only a few of the grammarians, I have expanded it by adding some long-pondered words of explanation. […] When I noticed that the grammarians had set forth conflicting rules, […] I thought it superfluous, even ridiculous, for me born of ignoble stock amongst the remotest tribes of Germania to pop up like a sheepish shepherd from behind a thorn bush or a clump of reeds and pronounce judgment. [… I]n each rule I have opted to follow that one found in holy treatises and in my daily reading to be the most often followed by the ecclesiastical dogmatists. [… N]ot so much as a single twig of a rule is grafted into this book without being firmly rooted in one of [Priscian, Donatus, Probus, Audax, Velius Longus, Romanus, Flavianus, Eutyches, Victorinus and Phocas]. [… O]nce you have scanned and understood all this [grammar], you will be able to contemplate the pages of Holy Scripture all the more lucidly. (Law 1997: 173f, slightly adapted) Boniface is acting as a ‘florilegist’ – collecting together the best excerpts from his authorities. In fact, Boniface did not use all the sources he claims to have used and he did use others unmentioned. 

The appeals to authority counter his being a non-native speaker of 138 The Western Classical Tradition in Linguistics Latin.15 Novelty is perilous. He rephrases his borrowings and substitutes his own examples to clarify the text. Following Charisius in Ars Grammatica (c. 350, Charisius 1964), he uses ordo for noun classes (declensions) and declinatio for case inflection. His intended readership was monks and nuns with some knowledge of Latin. Of course, he prefers the Latin of the Scriptures for his model, but unlike Smaragdus (writing about 805) who, as we have seen, commented on differences between grammarians’ prescriptions and Scripture, Boniface didn’t use examples from biblical texts. Only five of Boniface’s sources were the same as Tatwine’s: Donatus’ two Artes, Priscian’s De Nomine, Isidore’s Etym. Book I, and Asporius. Boniface also uses Charisius, Phocas, Audax (late classical period, Audax 1961), Diomedes (fourth century, Diomedes 1961), Sergius (Sergius 1961), Virgilius Maro, and Aldhelm. Alcuin or Alhwin (732–804) taught in his native York before joining Charlemagne16 in 781 as palace educator and librarian. He compiled the Epistula de Litteris Colendis (784–5) for Charlemagne, prescribing intensive study of Latin language and literature for all monastic and cathedral schools. In 796 he left the court in Aachen to become Abbot of St Martin in Tours, where he developed cursive minuscule script and promoted the illumination of manuscripts. Alcuin wrote treatises on grammar, orthography, rhetoric, and dialectic. His definition of grammar is not unlike that of Dionysius Thrax: ‘Grammar is knowledge of letters and knowledge of correct speaking and writing’ (Alcuin 1995: 857D). He was the first to exploit Priscian’s Institutiones Grammaticae (Minor) in his Dialogus Franconis et Saxonis de Octo Partibus Orationis “Dialogue between a Frank and a Saxon about the eight parts of speech” (Alcuin 1995), which presented grammatical instruction in easily digestible question-and-answer form. FRANK: Don’t forget that you said there were fifteen pronouns. Then why is it that Donatus included quis [“who”], qualis [“what sort”], talis [“such”], quot [“how many”], tot [“so many”], quotus [“what’s the number of”], totus [“all”] among the pronouns as well? SAXON: 

I remember saying that there were fifteen pronouns about which there could be no doubt. As for the ones you mention, there is doubt as to whether they are pronouns or nouns. Priscian, that ornament of Latin eloquence, says that they are interrogative, relative or redditive [rejoinder] nouns and says that they cannot be pronouns because they do not denote a definite person, which is one of the properties of pronouns that have case. (Alcuin 1995: 873C, see Law 1997: 137) Alcuin was a fulcrum for the so-called ‘carolingian renaissance’ because the influence of insular grammars spread from Charlemagne’s court. He trained the Italians Paulus Diaconus and Peter of Pisa who wrote insular-type parsing grammars based on Donatus; they were supplemented by paradigms from tracts such as Declinationes Nominum (unpublished) which was probably composed in early eighth century Britain (and often physically attached to the Ars Asporii). The Declinationes Nominum was a collection of noun paradigms using 15. Although Boniface wrote Latin with confidence, he spoke it with less confidence: when visiting the Pope in Rome he sought a written instead of an oral interview (Law 1997: 196). 16. Charles I (c. 742–814), King of the Franks and Lombards, and Holy Roman Emperor from 800. Prescriptivism from the early middle ages on 139 Priscian’s system. Paradigms were set out in running text to facilitate reading aloud. To save space any part thought predictable was omitted as in (1), a truncated form of (2): (1) hae uirtus tutis ti tem tus te tes tum bus tes tes bus [from a ninth century MS, cited by Law 2003: 135] (2) haec virtus virtutis virtuti virtutem virtus virtute virtutes virtutum virtutibus virtutes virtutes virtutibus Batches of such paradigms were copied from work to work with additions and subtractions resulting from scribal insights, error, preference, or whim; consequently, some versions of the Declinationes Nominum have more subtypes within first and second declensions than others. In one ninth century manuscript Jeudy finds copying errors in a list of diphthongs where the scribe had replaced AU with the word aut “or” (Jeudy 1993: 132). 

He also wrote ‘Hoc habent infinita, ideo non sunt aspirationes’ “Infinitives have this feature, and for this reason are not aspirations/aspirated”, thus misquoting Priscian’s non sunt separanda “are not to be separated [from verbs as parts of speech]” (ibid. 136; see also Law 1997: 21). Texts such as Iustitia Quid Est? “What is justice?” (Law 1982: 82), the eighth century Quae Sunt Quae? “What’s what?” (ibid. 86), and the eleventh century Beatus Quid Est? “What is holiness?” (Bayless 1993) present sets of questions exhausting every aspect of concepts raised and the terms used when discussing such words as beatus or iustitia; and they give model answers with commentary. Ælfric of Eynsham Anglo-Saxon England was the first country in Europe to develop and widely use a literary form of the vernacular, which it did from the time of the Benedictine reform c. 960. AngloSaxon glosses of Latin were common, but the structure of Old English was ignored. Literacy in Latin was dismal by the time that Ælfric (955–1010) of Eynsham in Oxfordshire wrote his grammar in the decade after 992. Ælfric’s Excerptiones de Arte Grammatica Anglice (Ælfric 11th c. manuscript; 1880; Porter 2002) used the vernacular as a medium for instruction, but in other respects it was a typical medieval elementary grammar of Latin. [O]nce you have studied the eight word classes of Donatus’ grammar in this book you will be able to incorporate both languages, Latin and English, into your tender minds until you arrive at more advanced studies. (Quoted in Law 1997: 207) So it was not training in Latinity, but a means of rendering Latin accessible to facilitate better understanding of the Scriptures. Ælfric presents lots of vocabulary in his subjectbased Glossary that was inspired by Isidore’s Etymologies. 17 [G]lossa, þæt is glesing, þonne man glesð þa earfoðan word mid eðran ledene “Glossa when one glosses the difficult words with easier words in Latin” (Quoted in Gneuss 1990: 18) 17. In Ælfric’s Homilies we find an Isidore-type etymology: Godspell is witodlice Godes sylfes lar “[the] gospel is certainly God’s own teaching” (Gneuss 1990: 25). 140 The Western Classical Tradition in Linguistics 

There was a conversation manual (the Colloquium), but, surprisingly, no prosody. Moreover, there are few or no insights into the grammar of either Latin or Old English. Latin definitions, examples, and quotations are translated using local colour and religious examples, for instance, uestri sutoris instrumenta – ?ower s?teres t?l “the tools of your cobbler”; Eadgarus Aðelwoldus rex episcopus “King Edgar, Bishop Athelwold”. Imperatiuus is beb?odendl?c, ac sw? ð?ah w? hit ?wendað oft t? gebede. Miserere mei, deus miltsa m?, god. “The imperative is for commands, but we use it often in praying: Have mercy on me, God.” (Quoted in Law 1997: 210) And there is exemplification-cum-commentary in Old English: Gif ð? cwest n?: hw? læ? rde ð?? Þonne cweðe ic: D?nst?n. Hw? h?dode ð?? H? m? h?dode. “If you were to say Who taught you? I would say Dunstan. Who ordained you? He ordained me.” (Law 1997: 208) The major source for Ælfric’s work is the extensive Excerptiones de Prisciano “Selections from Priscian” (Porter 2002), which, though structured on Donatus’ Ars Maior, takes its content from all of Donatus, Priscian’s Institutiones Grammaticae, Partitiones, and De Accentibus, and Isidore’s Etym. Book I. The Excerptiones clumps together material on the same topic that Priscian had distributed among different books. It is not a beginner’s grammar but serves as an introduction to or digest of Institutiones Grammaticae. Table 7.1. Ælfric’s parts of speech. LATIN ANGLO-SAXON declinatio nomen pronomen verbum adverbium participium coniunctio syllaba declinung, gebigednyss nama naman speliend word wordes gefera dælnimend geþeodnys, feging stæfgefeg Using loan-words, loan-formations, and semantic loans, Ælfric created an Old English metalanguage; there is a sample in Table 7.1. Many of these terms disappeared after the Norman Conquest. Ælfric expounds Latin morphology where he sees the necessity. For instance, for the Latin ablative he gives Preposition + Dative as the Old English equivalent. 

He draws attention to different gender assignments in the two languages (e.g. liber M, b?c F). He notes that the six Latin words et, que, ac, ast, at, atque all translate Old English and. He is thorough and not afraid to tackle difficult questions. ‘[In Britain] Ælfric’s grammar completely displaced other grammars at the elementary level soon after its appearance, and retained its dominating position into the twelfth century’ (Law 1997: 215). Prescriptivism from the early middle ages on 141 

The Anglo-Saxon grammar of Elizabeth Elstob Elizabeth Elstob (1683–1756) was a pioneer in studies of Anglo-Saxon, and the first woman whose grammar has survived (see above p.15). 

Her Rudiments of Grammar for the EnglishSaxon Tongue was based on ‘Ælfric’s Translation of Priscian’ and her grammar (Elstob 1715) is therefore very much in the Western Classical Tradition, though its influence was slight. On page 2 she distinguishes a linguistic sound, andgitfullic stemn [andgitfullic stemn], from the non-linguistic sound that an animal makes, gemencged stemn [gemencged stemn]. Then she turns to ‘letters’: A letter in Saxon stæf [stæf], is the least part of any Book or Writing, and cannot be divided. A Book or Writing may be divided into Words, S. cwydas [cwydas], those Words into parts, S. dælas [dælas], those Parts into Syllables, S. stæf gefgas [stæf gefgas], and afterward Syllables into stafas [stafas] Letters. Beyond this there is no further Division. In each Letter may be consider’d, its Name, S. Nama [nama], its Figure, or Shape, S. hiy [hiy], the same as our hue, its Power, S. miht[miht], i.e. what Power Letters have being join’d together with one another. (Elstob 1715: 2f) Elstob then discusses the five vowels, plus the ‘Greek’ vowel y ‘much used in Saxon’ (p. 3). The fricative consonants f, s, x, together with liquids and nasals, she describes as ‘semivowels’; the grouping may be partly explained by the fact that ‘the first six, ef, el, em, en, er, es, begin with the Letter e’. Stop consonants except for k are, of course, ‘mutes’. K is grouped with h and z because they respectively have the names ka, ha, za. Thorn (þ) and eth (ð) are not mentioned. 

Sounds are combined into syllables, syllables into words, and words into ‘speech or discourse’, but Elstob does not say how these larger categories are constructed. Giving Anglo-Saxon grammatical terminology throughout, Elstob identifies a traditional eight parts of speech, the first four of which decline: noun, pronoun, verb, participle, adverb, conjunction, preposition, and interjection. The noun has six cases (p. 7f); three genders; singular, plural and, occasionally, dual number. Following the Latin tradition of Ælfric, the article is included with nouns instead of being separated out as it would be in a Greek grammar; seven declensions are identified (pp. 10–12); simple and compound nouns are exemplified; adjectives, ‘Names gefera’, are described (pp. 17–19). Pronouns are subcategorized into singular, plural, and dual in the second person; primitive, possessive, relative, and reflexive pronouns are distinguished, and quantifiers such as some, a/one, all, along with cardinal numbers are included in the pronoun category. The verb (word) is defined as ‘a Part of Speech, with Time or Tense, and person, but without Case’ (p. 30; cf. Priscian 527 VIII.1, quoted above p. 116). Subcategorizations include active, passive, and neuter (intransitive); three tenses and six moods (indicative, imperative, optative, potential, subjunctive, infinitive). Irregularity is illustrated (pp. 45–48). The passive is discussed further in the section on participles (p. 43). The adverb is ‘always joined with a Verb and has not its full Signification without it’ (p. 49); Elstob lists 24 types. She also lists eight types of conjunction (p. 52). As in her models, Ælfric and Priscian, Elstob distinguishes those ‘prepositions’ (foresetnysse) which govern cases and those which are ‘used in 142 The Western Classical Tradition in Linguistics composition’ that today we call ‘prefixes’ (pp. 55f.). Finally she identifies eight classes of interjections. 

Under the heading ‘Of Syntax’ Elstob offers a brief discussion ‘Of the Construction and Ordering of NOUNS and VERBS’ (pp. 57–64) in phrases as well as sentences. There follow short accounts of the dialects of Anglo-Saxon, of Anglo-Saxon poetry, and an unsatisfactory mention of ‘accent’ (p. 69), which is said to distinguish e.g. God from gód “good” and man (indefinite pronoun) “one, people” from mán (neuter noun) “crime, sin” or (adjective) “bad, false”; Elstob fails to say that the accent marks a long vowel in such monosyllables. Elstob’s Rudiments of Grammar for the English-Saxon Tongue is at best a sketch grammar in the tradition of Dionysius Thrax and Donatus rather than Priscian. It is located in this chapter for that reason and because it was inspired by Ælfric’s grammar, rather than because it is a prescriptive grammar. Prescriptivism and standards in English grammar A descriptive grammar is one which sets out to describe the structure of a language, whether this is as an end itself or is to increase our knowledge of the language under description or of natural language in general. But any grammar which sets out to teach a language rather than to simply describe it must be prescriptive. Rules are given, but there is no philosophizing about them. Most of the grammars written early in the Classical Tradition had a primary pedagogical purpose. Dionysius Thrax and Apollonius Dyscolus described and exemplified good Attic Greek grammar and style for speakers of an Alexandrian koin?. Their Latin heirs, Donatus and Priscian, described good classical Latin for speakers of Late Latin or of Greek. Priscian created the prototype for parsing grammars in his Partitiones and he suggested some neat ways for learning new constructions on the basis of simpler constructions already known. Like their Alexandrian forebears, Donatus and Priscian disparaged solecisms, metaplasms,18 tautology, and verbosity. St Jerome’s Vulgate established Scripture as God’s own Latin, and ipso facto the best model that there could ever be; throughout the early middle ages and beyond, grammars in the Western Classical Tradition were Christianized. Tatwine, Boniface, and Alcuin all wrote competent Christianized Latin primers elaborating on Donatus’ Artes and Priscian’s De Nomine (plus, in Alcuin’s case, Priscianus Minor), with assistance from intervening exegetical works. In England, Ælfric of Eynsham’s Grammar and Colloquium were specifically intended as practical manuals for the teaching of Latin to speakers of Old English. 

In 1199 the Doctrinale (Alexander de Villa Dei 1893), which rapidly became popular all over Europe, used rhyme as a mnemonic to present a simplified version of Priscian’s grammar in Latin hexameters for students from non-Latin backgrounds who spoke laicae linguae. 19 The 18. A metaplasm is the alteration of a word by epenthesis, deletion or transposition of letters. Former U.S. President George W. Bush’s nukular/nucular is a metaplasm. 19. Other versifiers include Alexander Neckham (d. 1217) in Corrogationes Promethei; Eberhard Bethune’s Graecismus appeared about 1212; Ludolf of Luckowe’s Flores Grammaticae Prescriptivism from the early middle ages on 143 teaching method was similar to that of ancient Greece: starting with sounds, it moved on to vocabulary, parts of speech, some dialectic, rhetoric, and models of good writing; but all with a Christian motivation. Throughout, there was the heavy reliance on memorization, which is a consistent feature of pedagogical grammar. In 1540 William Lily (c. 1468–1522) wrote a grammar of Latin (Lily 1557; Lily and Colet 1970) that was prescribed by Henry VIII in 1542 for use in English schools. The ‘King’s grammar’ was traditional in favouring Virgil, Terence, and Cicero as models of good Latin; but came to be severely criticized because the rules for Latin were given in Latin rather than English. It was so well known that Shakespeare could refer to it as ‘the grammar’ in Titus Andronicus IV.ii when Chiron says, ‘O, ’tis a verse in Horace; I know it well: / I read it in the grammar long ago’, and obliquely in Henry IV, Part 1, ‘Homo is a common name to all men’ (Gadshill at the end of II.i.). English translations and commentaries were added by a number of seventeenth and eighteenth century grammarians. John Milton’s Latin grammar (Milton 1669) makes heavy use of it; and there are many others, some listed below. The long titles are given in full because they reveal the animosity and dogmatic self-confidence in the writer’s own preferred standard of ‘correctness’ contrasted with the fallaciousness of others. 

The Royal grammar reformed into a more easie method for the better understanding of the English, and a more speedy attainment of the Latin tongue (Wheeler 1695) A Treatise of the Genders of Latin Nouns: by way of examination of Lily’s Grammar rules, commonly called, Propria quae maribus; Being a specimen of Grammatical commentaries, intended to be published by way of subscription upon the whole grammar (Richard Johnson 1703) Grammatical Commentaries: being an apparatus to a new national grammar; by way of animadversion upon the falsities, obscurities, redundancies, and defects of Lily’s system. Necessary for such as would attain to the Latin (Richard Johnson 1706) which provoked a counter-attack Nolumus Lilium Defamari: or, A Vindication of the common Grammar, so far as it is misrepresented in the first Thirty Animadversions contain’d in Mr. Johnson’s Grammatical Commentaries. With remarks upon the same (Symes 1709) with a riposte Grammatical Commentaries: being an apparatus to a new national grammar; by way of animadversion upon the falsities, obscurities, redundancies, and defects of Lily’s system now in use. In which also many errors of the most eminent grammarians, […] are corrected […]. With an alphabetical index (Richard Johnson 1718). Lily’s text was revised in, e.g. A supplement to the English introduction of Lily’s Grammar: with select rules of the genders of nouns and the heteroclites. The whole from Lily’s Latin grammar, publish’d at Oxford. For the use of the school in Exon, commonly call’d the Free-School (Lily 1719) appeared mid-thirteenth century; John of Garland (d. c. 1272) wrote Compendium Grammatice and Clavis Compendii (see Law 1997: 267). 144 The Western Classical Tradition in Linguistics Propria Quae Maribus: quae genus, as in praesenti, syntaxis, qui mihi, construed. Printed according to act of Parliament. Sold by all the booksellers in town and country. (Lily, Robertson and Ward 1776) Lily’s grammar was in use for more than three hundred years. 

The eighteenth century grammarians of English were mainly clergymen like the Reverend Dr Robert Lowth (1710–87), retired gentlemen like Robert Baker (fl. 1770), and amateur philosophers like George Campbell (1719–96). One wrote for the ‘Female Teacher in the British Dominions’ (preface to Farro 1754: vii); some wrote for the improvement of persons in trade or manufacturing; but ‘the majority of writers seem to have felt that they were writing for the edification and use of gentlemen, to warn them against inadvertent contamination with the language of the vulgar’ (Leonard 1929: 169): Purity and Politness of Expression […] is the only external Distinction which remains between a Gentleman and a Valet; a Lady and a Mantua-maker. (Withers 1789: 161) Although this classism was the norm, it was not universal. If a Man were with a serious Countenance to ask a Servant-Wench, that is standing at a Door, what a Noun Adjective is, and whether such a Verb governs a Dative or an Accusative Case, she would conclude him to be out of his Senses; and would perhaps run frightened into the House, and tell her Mistress that a Madman was going to do her a Mischief. And yet this Wench, who never heard a Word of Prepositions, Participles, Substantives and Verbs, makes use of all these Parts of Speech (and, generally speaking, very properly) without knowing they have any Names. (Baker 1770: v) The prevailing attitude of the eighteenth century was put by Jonathan Swift (1667–1745) in A Proposal for Correcting, Improving, and Ascertaining the English tongue. I do here, in the Name of all the Learned and polite Persons of the Nation, complain [...] that our Language is extremely imperfect; that its daily Improvements are by no means in proportion to its daily Corruptions, that the Pretenders to polish and refine it, have chiefly multiplied Abuses and Absurdities, and that in many instances, it offends against every part of Grammar. (Swift 1712: 8) Dr Samuel Johnson (1709–84) wrote in the Preface to his Dictionary: ‘I found our speech copious without order, and energetic without rules’ (Johnson 1755). 

Views like those of Swift and Johnson led to an outflow of prescription for what was proper in English. Swift gave as one reason for the inferiority of English that ‘the Latin tongue in its purity was never in the Island’ (op.cit. 9). The supposed superiority of Latin was a constant theme, though Greek was elevated to a higher position, with Latin universally classified ‘a Species of Greek somewhat debased’ (James Harris 1786: 148). Romance languages were vulgar corruptions of Latin; and Germanic languages were barbaric to most except Elstob 1715. Buchanan 1767: ix, finding errors in the work of Swift, Addison, and Pope, wrote ‘Had they not the Rules of Latin Syntax to direct them?’ Lowth 1763: 42 (Lowth 1789: 47) wrote that ‘the Double Superlative most highest is a Phrase peculiar to the Old Vulgar Translation of the Psalms; where it acquires a singular propriety from the Subject to which it is applied, the Supreme Being, who is higher than the highest’; for which he was criticized by New York maths teacher Mennye 1785 on the basis that the Latin translation ‘maxime altissimus would Prescriptivism from the early middle ages on 145 be ridiculous’. George Harris 1753: 20f says like ‘ought never to be used when it cannot be translated into Latin by the Word Similis.’ There was, of course, a contrary view: I am not ignorant, that the Practice may be supported by the Syntax of ancient Languages. But what have we to do with foreign Idioms? It is Wisdom to enrich our Vocabulary with Words from every Quarter of the Globe; but an Indignity to suffer any Nation to controul our Style. (Withers 1789: 40f) It was rare for Latin to be rejected as a model. Blair put a compromise position, which relied on general or universal grammar (see Chapter 8): ‘All the rules of Latin syntax, it is true, cannot be applied to our language. [… But] the chief and fundamental rules of syntax are common’ (Blair 1783 I: 166f ). And Withers 1789: 150f notes that Latin persecutio means “prosecution” and that prosecutio means “persecution”, so recourse to Latin when seeking the meanings of English words can lead to error.20 Noah Webster (1758–1843), favouring description over prescription, recognized that grammars should show ‘what a language is – not, how it ought to be’ (Webster 1789: 204ff) and is very modern when he writes ‘spoken language […] is the only true foundation of grammar’ (Webster 1798: 15f); nevertheless, he appealed to Latin when discussing whether or not the preposition should accompany the interrogative pronoun in wh- questions, and what form the wh- word should take, and therefore was convinced that ‘whom do you speak to? is a corruption’, preferring who (Webster 1789: 287).21 Appealing to reputable literary authorities to set the standard was a problem. I have censured even our best Penmen, where they have departed from what I conceive to be the Idiom of the Tongue, or where I have thought they violate Grammar without necessity. (Baker 1770: iv) The classical English authors were all criticized for solecisms by some grammarian or other. Shakespeare, because of Ben Jonson’s evaluation of his ‘little learning’, was expected to use vulgarisms. The ‘wrong’ use of prepositions even by Swift, Temple, Addison, and other writers of the highest reputation; some of them, indeed, with such shameful impropriety as one must think must shock every English ear, and almost induce the reader to suppose the writers to be foreigners. (Baker 1770: 109) Lowth 1762: iii, among others, later reported much the same. As we have seen, there was often acrimony involved, since what was correct to one writer was incorrect to another. ‘Lowth censures Addison for writing “as either of these two qualities are wanting.” […] Priestley censures, in Smollett’s Voltaire “The number of inhabitants were not more than …”’ (Leonard 1929: 218, 220). Webster 1807: 16f accuses Dr Johnson of using attain in the sense obtain. Johnson 1755 writes of flee ‘This word is now almost universally written fly, 20. The lesson has some force, but Withers is not altogether accurate: accusatio is more frequently “prosecution”, and prosequor does not normally mean “persecute”; insectatio is “persecution”. 21. Murray 1795: 122 wrote, ‘the placing of the preposition before the relative [pronoun] is more graceful, as well as more perspicuous, and agrees much better with the solemn and elevated style’; but not in conversation or informal writing. For more on preposition stranding see Yáñez-Bouza 2008. 146 The Western Classical Tradition in Linguistics although properly to fly […] is to move with wings and flee […] to run away.’ Lowth 1789: 118f objects In a few instances the Active Present Participle hath been vulgarly used in a Passive Sense: as beholding for beholden: owing for owen … “I would not be beholding to fortune for any part of the victory.” And he says that, contrary to what was often heard, whereas I mistake/am mistaking means “I misunderstand”, I am mistaken means “I am misunderstood”. Baker 1770: 11 complains ‘To Demean signifies to behave, to comport, and not to debase or lessen.’ 

Elsewhere he prescribes ‘I get, go, or come up or down stairs; I am above or below stairs’ (ibid. 15). On p. 7f he favours different from over different to, though he admits the latter is frequently used, even by ‘good Writers’; he did, however, reject different than altogether. Many authors used you was for the singular, but Lowth 1789: 55 dismisses it as ‘an enormous solecism’; Campbell 1776, I: 339ff says you was is colloquial, like there’s the books. Both these expressions persist today and are condemned by purists – whose past strictures have obviously been ineffective. There was much discussion of Mark 8: 27 (King James version) Whom do men say that I am? The almost universal view was that the translators erred: ‘Whom’ should be Who. 22 Murray 1795: 113, for instance, points out that the interrogative pronoun is governed by ‘am’, not ‘say’, and should agree in nominative case with I. A lot of the argument was concerned with what are surely stylistic and dialect variations. Particularly blatant examples of pure prejudiced personal preference are to be found in J. Johnson 1763: 19, who castigates as ‘Scotticisms’ a pretty enough girl instead of a pretty girl enough; paper, pen and ink instead of pen, ink and paper; and nothing else for no other thing. Baker 1770: 115 had written, ‘These seeming Minuties are by no Means to be despised, since they contribute to the Intelligibleness of Language’ – a valid point, but it does not apply to J. Johnson’s reviled ‘Scotticisms’. Ann Fisher was ‘the first woman to produce a grammar of English’ (Tieken-Boon van Ostade 2008a: 145f); moreover, as was remarked in Chapter 1 (p. 16), her grammar was, deservedly, extremely popular in the late eighteenth and early nineteenth centuries. Fisher believed in a universal grammar such that study of the grammar of any one language would facilitate command of another (see Chapter 8 below). A Person therefore who understands English grammatically, must be allowed to have good Notions of Grammar in general; i.e. that of every other Nation, and consequently, if he endeavours to learn any other Tongue, will, from this Analogy, find his Progress surprisingly facilitated. (Fisher 1753: ii [sic]) [T]he Reason why those among us, who have learned Latin, &c. are greater Adepts in our own Language than those who have only learned English at Random, or ingrammatically, is entirely from their Knowledge of Grammar in general; which they acquire by learning such or such Languages by It: For though every Language has its peculiar Properties or Idiotisms, the Nature of GRAMMAR is, in a great Measure, the same in all Tongues. (ibid. iv [sic]) Q. Are the Parts of Speech the same in English as in Latin? 22. As it is in other translations, see http://bible.cc/mark/8-27.htm. Prescriptivism from the early middle ages on 147 A. Yes, and in all other Languages as well as Latin: For that which is a Name, or Noun Substantive in English, is a Noun Substantive in the Latin, Greek, Hebrew, French, &c. Languages. (ibid. 61) Fisher was convinced that English is at least the equal of, if not superior to, any other language. 

That the English Language is as copious, significant and harmonious as any in the World, none pretend to dispute. (ibid. ii) If to be Master of any Language, so as to write it with Propriety and Exactness, is to understand it grammatically; it must certainly be a nearer or more concise Way, to the Perfection of ours, to learn the English Grammar itself, than to go about to study the Latin One, &c. merely to come at the Knowledge of our own from the Nature of theirs. (ibid. v) As in English we have but this one Case [the Genitive Case, or Possessive Name], we express the Circumstances, Properties, or Affections of Things to one another by the Help of little Words called Prepositions, such are of, to, with, from, by, &c. whereby we are freed from the great Trouble that is found in other Languages of expressing the Circumstances, &c. of Names in twelve Cases, and five or six different Declensions: So likewise our having no difference of Gender in our Names, is an Advantage as great as the former, and which no other Language Antient or modern enjoys, except the Chinese. (ibid. 70) Fisher’s New Grammar with Exercises of Bad English has four parts (i) Orthography, dealing with spelling, syllabification, and punctuation; (ii) Prosody, pronunciation; (iii) Etymology, parts of speech and their morphology (‘so very essential to polite Writing, that I cannot think any one qualified to speak, write, or compose with a happy Propriety, a Clearness, and Comprehensiveness of Expression, who has not a thorough Knowledge of and Regard to it’ (ibid. x); (iv) Syntax, ‘which shews how to connect Words aright in a Sentence, or Sentences’ (title page). The method of instruction is question and answer, like that of Donatus (see Chapter 5), e.g. Q. What is Grammar? A. Grammar is the Art of expressing the Relation of Things in Construction, with due Accent in speaking, and Orthography in Writing, according to the Custom of those whose Language we learn. (Fisher 1753: 1) Part I on Orthography and Part II on Prosody are thought to have been the particular interest and contribution of Daniel Fisher in the first edition of the book;23 but if that was indeed so, Ann Fisher must have been an apt pupil. In any case, it seems clear that these two parts also owe a great debt to Thomas Dyche (d. c. 1733), see Dyche 1707.24 

The section on Orthography deals with letters and their pronunciation in words. Vowels and consonants are defined. Vowel length and its distribution in English are discussed; there is no mention of the term monophthong, only of diphthongs and triphthongs. The focus on written language is evident in that an ‘Improper Diphthong’ is one like the eo in people pronounced as ‘e long’; puzzlingly, another is the double aa in Aaron. Fisher deals with y and w as each being both vowel and consonant; she examines some of the vagaries of English spelling with 23. See above p. 16. 24. The Fishers may have used the later 24th edition, corrected, published in London by Richard Ware in 1737. 148 The Western Classical Tradition in Linguistics respect to pronunciation, and gives the correct pronunciation for borrowed Greek, Hebrew, Latin, and French words (pp. 14, 18, 24). Her description of syllables in English is thoughtful and apt; it has rules of syllabification which include some remarks on derivational morphology (p. 34) – which is also written of under Etymology in Part III. Punctuation is discussed, which is beyond the interest of a linguistics student, but not that of the English language student for whom this grammar is written. In Part II Prosody is explained. Word stress distribution is examined in polysyllabic words. The different stress marking of noun and verb in words like accent, concert, desert, etc. are identified (e.g. áccent vs accént). A term from Dyche 1707, Double Accent, is defined as follows: [T]he Use of [? the Double Accent] is every where to denote, that the Letter which begins the Syllable to which it is prefixed has a double Sound, one of which belongs to the preceding Syllable: Thus the Words Ba?-lance, Cha?-pel, Mi?-stress, A?-ni-mal, &c. are sounded with double Consonants; as, Bal-lance, Chap-pel, Mis-stress, An-ni-mal. (Fisher 1753: 47) We see that pronunciation has changed since the mid-eighteenth century; this is also evident for certain other pronunciations given in Fisher’s phonetic spelling such as perfect (perfit), perfected (perfited), Chorister (Quirister), Ribband (Ribbin) (ibid. 18f). Part III is on Etymology ‘or, the kinds of words’. Etymology was not interpreted like etymologia among the ancients and early medievals, but as morphology or word-andparadigm grammar, as the term etymologia was used by the modistae in the late middle ages (see Chapter 8). Part III deals with derivational morphology and the four parts of speech: Names, ‘which express Things, or Substances; Qualities, ‘which express the Manners, properties, or Affections of Things’; Verbs, ‘which express the Actions, Passions, or Beings of Things’; Particles ‘shew the Manner or Quality of Actions, Passions, or Beings, &c.’ Names are subcategorized into common, proper, and ‘Relative Names, or Pronouns’. The latter are different from other Names in having a form for the ‘leading State’ (subject form, e.g. I) and ‘following State’ (oblique form, me). Number, gender, and the genitive case are illustrated. Qualities (adjectives) can be identified ‘By putting the Word Thing after them, which they will bear with good Sense; as, a good Thing, … [A] Quality cannot clearly signify any Thing, without a Name either expressed or understood; as, […] refuse the evil (Thing) and chuse the good: Thing is, in both places, understood’ (p. 71). What today would be called determiners are included among Qualities; so too participles, as in a scolding Woman, a ruined Man. The Verb ‘is that Part of Speech which betokens the doing, being, or suffering of a Thing; to which belongs the several Circumstances of Person, Number, and Time’ (pp. 79f). Auxiliaries are named ‘Helping Verbs’. 

Four kinds of Particles are identified: Adverbs, Conjunctions, Prepositions, and Interjections. In traditional fashion Prepositions divide into those ‘set separate or before other Parts [… or] joined or set in composition’ (p. 96), the latter being prefixes. Discussing the derivation of words, conversion is described, e.g. ‘from a House comes the Verb to house (houze)’ and ‘almost every Verb has some Name coming from it; and, by adding the Termination er to a Verb comes a Name, signifying the Agent or Doer’ (p. 103). Many other derivations are also identified. Because this is a pedagogical grammar there are parsing exercises (e.g. p. 109). Prescriptivism from the early middle ages on 149 Part IV is about Syntax, or Construction, ‘The right joining of Words in a Sentence, or Sentences together’ (p. 112). Just ten rules of syntax are offered (pp. 113–17). Some perceived exceptions ‘being authorised by Custom, and not reducible to Rule, may be called Anglicisms, viz. a few Days; many a Time; me thinks; every ten Years; whilst the Book was a-printing; whilst the Stream was a running, &c.’ (p. 121). Fisher discusses normal and unusual word order in sentences with the urging to ‘follow the Use of the best Speakers and Writers’ (p. 123). Pp. 126ff offer many examples of bad English with correction exercises; misspellings in these give some clue as to contemporary colloquial pronunciation, for example ‘youmer’ for humour, ‘featers’ for features, ‘natral’ for natural, ‘pictors’ for pictures. Part of the advice on reading aloud is that different meanings result from different sentence stress, e.g. ‘there may possibly be four different Senses, from the different placing of Emphasis’ on Will you ride to the Town To-day and these are elaborated (p. 148). Fisher’s New Grammar is an excellent example of a pedagogical grammar of English; if it makes no original contribution to grammatical theory, this is typical of its genre. 

She disagreed with the Dean of St Patrick’s and many others that English was inferior to Latin and that it was intrinsically imperfect; instead she dwelt on how proper instruction in English would establish grammatical usage among its speakers and render the English language as copious, significant and harmonious as any in the world. Lowth’s A Short Introduction to English Grammar was also written in response to Swift. Like Fisher, Lowth agreed that many people speak and write ungrammatically, but he did not believe that English is by ‘nature irregular and capricious’ and he set about describing for it ‘a System of rules’. His grammar has a very traditional format, discussing in turn letters (and their pronunciation), syllables, words, and nine parts of speech: article, noun, pronoun, adjective, verb, adverb, preposition, conjunction, and interjection. His grammar also had sections on sentences, punctuation, and gave exercises in parsing. The principal design of a Grammar of any Language is to teach us to express ourselves with propriety in that Language, and to be able to judge of every phrase and form of construction, whether it be right or not. […] A good foundation in the General Principles of Grammar is in the first place necessary for all those who are initiated in a learned education; and for all others likewise who shall have occasion to furnish themselves with the knowledge of modern languages. […] When [a person] has a competent knowledge of the main principles, the common terms, the general rules, the whole subject and business of Grammar, exemplified in his own Language; he then will apply himself with great advantage to any foreign language, whether ancient or modern. [… A] competent Grammatical knowledge of our own Language is the true foundation upon which all Literature, properly so called, ought to be raised. (Lowth 1762: x–xii) Lowth states that strong verbs like write and ride should distinguish between past tense and past participle forms, identifying what he regarded as ‘common mistakes’: He begun, for he began; he run, for he ran; he drunk, for he drank: The Participle being used instead of the Past Time. And much more frequently the Past Time instead of the Participle: as, I had wrote, it was wrote, for I had written, it was written; I have drank, for I have drunk; bore, for born; chose, for chosen; bid, for bidden; got for gotten &c. This abuse has been long growing upon us, and is continually making further incroachments. (Lowth 1762: 86–89) 150 The Western Classical Tradition in Linguistics In his own private informal correspondence, however, Lowth flouted this grammatical rule. In a letter to his wife we find ‘My Last was wrote in a great hurry’; and later in the same letter ‘whose faces and names I have forgot’. We might accuse Lowth of hypocrisy because he was willing to violate his own theoretical prescriptions when writing informally to his wife or to close friends, but he took care not to make the same ‘mistakes’ in more formal letters (Tieken-Boon van Ostade 2002: 463–5). People have confused past tense and past participle forms of strong verbs since the beginning of the medieval period (Lass 1999: 166ff); language is not amenable to being forced into a standard mould and anyone who attempts to do so will find themselves bemired in contradiction. American-born Lindley Murray (1745–1826) settled in England in 1784. His Grammar is very similar in design to Lowth’s but is more comprehensive and better laid out, using bigger print for more important points and numerous sets of rules and verb paradigms (a teaching method indebted to Claude Lancelot’s instructional grammars of several languages, e.g. Lancelot 1681; 1758 – the latter an English translation). Murray’s book had four main sections. Orthography dealt with letters and their pronunciation, syllables, and spelling. 

As for Fisher 1753, ‘Etymology […] treats of the different sorts of words [nine parts of speech], their derivation, and the various modifications by which the sense of a primitive word is diversified [by subcategories such as number, case, tense, mood]’ (Murray 1795: 19). His definitions of parts of speech are a mix of formal and semantic criteria, for example, An article is a word prefixed to substantives, to point them out and show how far their signification extends. [...] A substantive or noun is the name of anything that exists or of which we have any notion. It may in general be distinguished by taking an article before it or by its making. (ibid. 41) [A] word which has the article before it, and the possessive preposition of after it, must be a noun. (ibid. 117) Murray floats the idea that there are as many cases in English as ‘the various combinations of the article and different prepositions with the noun’ (ibid. 28), but he doesn’t follow it up; however, he does see the need to exemplify English cases via Latin: Nominative. MAGISTER, A master. Genitive. MAGISTRI, Of a master. Dative. MAGISTRO, To a master. Accusative. MAGISTRUM, The master. Vocative. MAGISTER, O master. Ablative. MAGISTRO, From or by a master. (ibid. 26) Murray should have followed Fisher’s better example (see the quote p.147 above from Fisher 1753: 70). Murray uses a transitive clause to identify ‘the objective case’ (ibid. 29) ‘Syntax […] shews the agreement and right disposition of words in a sentence’ (ibid. 86). He discusses the parts of speech in the sequence article, noun, pronoun, adjective, verb, adverb, preposition, conjunction (including here comparatives), and interjection. In the Stoic tradition he claims that ‘Two negatives, in English, destroy one another, or are equivalent to an affirmative’ (ibid. 121) but as Cameron so rightly says: Prescriptivism from the early middle ages on 151 I have yet to meet any speaker of any variety of English who on hearing Mick Jagger sing ‘I can’t get no satisfaction’ has entertained for one moment the belief he means the opposite. (Cameron 1995: 25)25 The same was true in the eighteenth century. 

Murray includes more parsing exercises than Lowth. His discussion of prosody reviews ‘accent’ (stress), ‘quantity’ (syllable length), ‘emphasis’ (sentence stress), ‘cadence’ (downdrift intonation), feet in versification, and punctuation: The Comma represents the shortest pause; the Semicolon, a pause double that of the comma; the Colon, double that of the semicolon; and the Period, double that of the colon. (Murray 1795: 159) It is a moot point whether this is accurate even for texts read aloud; it certainly needs revising in terms of degree of semantic-syntactic disjuncture for written texts. Murray is very modern in his recommendations for the use of capital letters. The Grammar ends with an Appendix containing Rules and Observations for promoting perspicuity and accuracy in writing. This is a precursor to the Gricean maxims (Grice 1975) and is closely modelled on John Hughes’ essay ‘Of Style’, written in 1698 and circulated thereafter (Hughes 1915). Like Hughes, Murray favours ‘purity’ (avoid obsolete, new-coined, ungrammatical, or foreign expressions), ‘propriety’ (avoid violations of what would come to be called the manner maxim), and ‘precision’ (choose the right expression and observe the quantity maxim); sentences should be clear, relevant, and have internal unity. He champions semantic extension and figurative usage because they ‘enrich language […] give us, frequently, a much clearer and more striking view of the principal object’ (Murray 1795: 212) and because they are necessary: ‘No language is so copious, as to have a separate word for every separate idea’ (ibid. 211). Prescriptive he may have been, but there is much good sense in Murray 1795. In a link back to the early middle ages, Holy Writ was his standard: In the course of this work [his Grammar], some examples will appear of erroneous translations from the Holy Scriptures, with respect to grammatical construction; but it may be proper to remark, that not withstanding these verbal mistakes, the Bible, for the size of it, is the most accurate grammatical composition which we have in the English language. […] Dr Lowth […] says, “The present translation of the Bible is the best standard of the English language.” (Murray 1795: 103) He deliberately used uplifting examples such as Blot out all mine iniquities and The man is happy who lives virtuously (p. 32); The scholars were attentive, industrious, and obedient to their tutors; and by these means acquired knowledge (p. 102). His Grammar was popular throughout the nineteenth century and it was supplemented by exercises (Murray 1799). Aristotle’s observation that language is conventional was generally accepted from his day on. John Locke (1632–1704) in his Essay Concerning Humane Understanding had reiterated the notion, disclaiming any inherent or necessary link between a word and its denotatum (Locke 1700: III.ii.1). Locke was an empiricist – like his friends and fellow members of the Royal Society, Robert Boyle and Isaac Newton. The Essay was to be ‘a 25. As pointed out more prosaically in Chapter 4 the double negative in English I cannot not go is synonymous with I must go, which implies, but is not implied by, I can go. 152 The Western Classical Tradition in Linguistics Survey of our own [human] Understandings, examine our own Powers, and see to what Things they were adapted’ (ibid. I.i.7). ‘Ideas’ provide the mind with representations of objective qualities of objects (such as size, shape, or weight) and also secondary qualities such as colour, taste, or smell which are subjective (ibid. Book II). To understand thinking and knowing one must understand language as the means of thought and communication (Book III). Locke claimed that words only mean what they are understood to mean; consequently, usage must be the sole arbiter. Words in their primary or immediate Signification, stand for nothing, but the Ideas in the mind of him that uses them. (ibid. III.ii.2) Linguistic forms represent the ideas of things and not the things themselves (ibid. III.ii.5). Frequently, the idea signified is not clear, and sometimes words are used even when there are no ideas corresponding to them. This is particularly so in the case of generic terms and universals. Locke suggests that such terms are nothing more than a creation of the mind, through abstraction; they denote ‘nominal essences’. 

He also discusses definitions, and rejects the Aristotelian view that all definition must be per genus et differentiam, saying that this only works so long as the Idea behind a word is explained: For Definition being nothing but making another understand by Words, what Idea, the term defined stands for, a definition is best made by enumerating those simple Ideas that are combined in the signification of the term Defined. (ibid. III.iii.10) Locke identifies three criteria for efficient communication: First, To make known one Man’s Thoughts or Ideas to another. Secondly, To do it with as much ease and quickness, as is possible; and Thirdly, Thereby to convey the Knowledge of Things. Language is either abused, or deficient, when it fails in any of these Three. (ibid. III.x.23) Locke’s embrace of the criterion of usage was less prestigious among the eighteenth century commentators on grammar than the views of Horace and Quintilian. Horace wrote, c. 13 BCE, ‘the will of custom, in the power of whose judgment is the law and the standard of language’ (Ars Poetica 71f, Horace 1928). Quintilian similarly: Usage, however, is the surest pilot in speaking, and we should treat language as currency minted with the public stamp. (Quintilian 1920-22, I.vi.1–3) [W]e must make up our minds what we mean by usage. If it be defined merely as the practice of the majority, we shall have a very dangerous rule affecting not merely style in language but life as well, a far more serious matter. For where is so much good to be found that what is right should please the majority? [...] I will therefore define usage in speech as the agreed practice of the educated. (ibid. 43–45) Eighteenth century grammarians would doubtless claim to be presenting ‘the agreed practice of the educated’ in their works; but in fact there seems to have been more dispute than consensus over what constitutes good usage: each grammarian presented his own judgments which, as we have seen, often disagreed with those of his fellows. Any appeal to good usage ought to describe what is meant by such a phrase; and the better grammarians did so. For instance, George Campbell in The Philosophy of Rhetoric identified it as ‘reputable, national, and present’ (Leonard 1929: 148). By ‘reputable’ he means used at Court or found in authors acknowledged by educated gentlemen (Campbell 1776, I: 351f). By ‘national’ he Prescriptivism from the early middle ages on 153 means neither provincial nor foreign (ibid. 353). 

By ‘present’ he means within the previous century, though he apparently excludes examples of good usage from living authors (ibid. 366). We have already seen that an appeal to the standards set by authors of reputation was fraught with controversy over their perceived occasional ungrammatical usage. Campbell’s canons of good usage (Leonard 1929: 154–60) include: (a) unambiguous expression; (b) regular rather than irregular forms; (c) simplicity rather than complexity; (d) euphony; (e) conformity with Latin or Ancient Greek syntax; (f) avoiding solecisms, barbarisms, etc. (a), (c) and (f) seem very reasonable by today’s standards. Even (d) is a reasonable criterion for good style. But (e) we would now reject; the eighteenth century views on the matter have already been discussed. (b) is a quaint preference for ‘analogy’ such as led George Harris 1753: 24 to be sad that the regularizations knowed, falled, rised cannot be (re)introduced; in the light of (e), though, it is doubtful whether Campbell would have approved such regularized plurals as syllabuses and phenomenons. The rational justification for an appeal to the classical languages as models for English grammar was a belief in ‘universal’ or ‘general’ grammar:26 ‘the chief and fundamental rules of syntax are common’ to all languages (Blair 1783, I: 166f ). All men, even the lowest, can speak their Mother-tongue. Yet how many of this multitude can neither write, nor even read? How many of those, who are thus far literate, know nothing of that Grammar, which respects the Genius of their own Language? How few then must be those, who know GRAMMAR UNIVERSAL; that Grammar, which without regarding the several Idioms of particular Languages, only respects those Principles, that are essential to them all. (James Harris 1786: 11) On the presumption that so-called ‘modern’ languages are corruptions of the ancient languages, the latter were supposedly closer to general grammar. 

Harris used many Greek and Latin examples. His universal grammar admitted parts of speech and other syntactic phenomena from familiar languages but excluded the unfamiliar, such as postpositions (which occur in, for example, Hungarian, Japanese, and Turkish). He claimed that the sun and moon are naturally masculine and feminine respectively – in defiance or ignorance of the Germanic and Slavic languages. For such infelicities he was rightly condemned by John Horne Tooke 1786. Notice the implicit assumption that the earliest language was closest to, if it did not embody, general grammar. They [eighteenth century grammarians] built in general upon the neo-Platonic notion of a divinely-instituted language, perfectly mirroring actuality but debased by man, and they labored to restore its pristine perfection. (Leonard 1929: 14) One pervading view was that grammar mirrors nature: 26. 

The term ‘general’ in general grammar derives from Latin genus~generis; the unique aspects of grammars of particular languages manifest species in their accidentia. The history and development of general grammar is discussed in Chapter 8. 154 The Western Classical Tradition in Linguistics THOSE PARTS OF SPEECH UNITE OF THEMSELVES IN GRAMMAR, WHOSE ORIGINAL ARCHETYPES UNITE OF THEMSELVES IN NATURE. (Harris 1786: 263f [sic]) This is a hangover from the speculative grammars of the late middle ages (discussed in Chapter 8). Similar was Burnet 1773-92 Vol II: 343 and Withers 1789: 124 – who, in this context, described Nature as ‘the Source of real Elegance’ and then explained ‘what are the Laws of Nature but the Operations of the Deity?’ (ibid. 185f). Those eighteenth century grammarians who championed usage as a criterion of grammaticality were usually contrasting it with a notion of general grammar. This occurs, for example, in Campbell’s ridicule of Swift for thinking that language could offend ‘against every part of Grammar’ (Swift 1712: 8, quoted above p. 144). What could the Doctor’s notion of grammar be, when he expressed himself in this manner? Some notion, possibly, he had of grammar in the abstract, an universal archetype by which the particular grammars of all different tongues ought to be regulated [...] I acknowledge myself to be entirely ignorant of this universal grammar; nor can I form a conjecture where its laws are to be learnt. [...] One thing indeed every smatterer in philosophy will tell us, that there can be no natural connection between the sounds of any language, and the things signified, or between the modes of inflection and combination, and the relations they are intended to express. [...] If he meant the English grammar, I would ask, whence has that grammar derived its laws? If from general use (and I cannot conceive another origin), then it must be owned, that there is a general use in that language as well as in others; and it were absurd to accuse the language, which is purely what is conformable to general use in speaking and writing, as offending against general use. But if he meant to say, that there is no fixed, established, or general use in the language, that it is quite irregular, [...] he ought to have said, that it is not susceptible of grammar; which, by the way, would not have been true of English, or indeed of any of the most uncultivated languages on the earth. (Campbell 1776, I: 342f, quoted in Leonard 1929: 149) 

There was another view of language, too, which stemmed from rationalist grammar (also discussed in Chapter 8). This used the principle that language is conventional to argue for improving language by rational means to modify linguistic conventions. It permits the quibble that if a form or expression is logical, then it must be acceptable; if it is illogical it is unacceptable. What was logical was general grammar, which, as we have just seen, was perceived to be most closely represented by the grammar of a classical language. A number of eighteenth century grammarians appealed to the ‘genius’ of the language, e.g. Harris 1786: 11, quoted above, and Johann Michaelis: Every one who is master of the language he speaks [...] may form new words and form new phrases, provided they coincide with the genius of the language. (Michaelis 1771: 79) The term genius of the language27 retains some of the classical sense of genius as a tutelary god or controlling spirit of the language which determines its distinctive character; but it had come to mean what the native speaker of the language (perhaps only the educated gentleman) intuitively finds correct (Leonard 1929: 29). Certainly, the mystical quality of 27. The earliest reference I know of is Petit 1686: 116: ‘Les Romains entendoient leur langue avec la méme facilité que nous entendons la nostre. Chaque nation a disposé la sienne selon son genie, sans songer si l’on garderoit le mesme ordre dans les autres.’ Prescriptivism from the early middle ages on 155 ‘genius’ comes out equivalent to a grammarian’s opinion of what best matched the grammatical norms of the language. Thus, as the grammarian established the grammar of the language, he concomitantly identified the ‘genius’ of the language, to which he could then appeal when justifying his pronouncements. There was, therefore, a vicious circularity. The eighteenth century prescriptivists had very few new insights into linguistic analysis. There were some, however. For instance, Hugh Blair, writing of the omission of the relative pronoun (in such a phrase as the man I loved), notes, But though this elliptical style be intelligible, and is allowable in conversation and epistolary writing, yet, in all writing of a serious or dignified kind, it is ungraceful. 

There, the relative should always be inserted in its proper place, and the construction filled up. (Blair 1783, I: 221f, 453) This shows a keen awareness of the difference between language usage in the spoken and the written medium – essentially a difference in the style of language used. On the other hand, Withers 1789: 405 thinks that omission of the relative pronoun is acceptable in writing. Fisher 1753 and Priestley 1761 noted that English has no future (syntactic) tense, something that is not always recognized even today. Much eighteenth century linguistic criticism was, however, oriented to matters of style. John Hughes’ essay ‘Of Style, written at the Request of a Friend in the Year MDCXCVIII’ (so, a little before the century began) has some ideas that one may compare with one of Grice’s maxims of the cooperative principle (Grice 1975), namely, manner: All the Qualifications of good Style I think may be reduced under these four Heads, Propriety, Perspicuity, Elegance, and Cadence: And each of these except the last, has some relation to the Thoughts, as well as to the Words. [...] There is another Particular which I shall mention here, [...] and that is Purity, which I take more particularly to respect the Language, as it is now spoke or written. (Hughes 1915: 80f) Blair (1793, I: 439) takes perspicuity to be ‘the fundamental quality of style’, with purity, propriety, and precision as its subcategories. He says the necessity for precision is ‘not only that every hearer may understand us, but that it shall be impossible for him not to understand us’ (1793, I: 171). As noted already, this was taken up by Murray 1795: 177ff. Most of the discussion on style in the eighteenth century seems trivial to a modern linguist. The usual procedure was to complain of lack of clarity, logicality, or propriety on grounds that are almost invariably based on the personal belief and subjective judgment of the grammarian. The emphasis was on the ‘correctness’ of each word, while clarity or force in the communication of ideas was almost ignored. The concern with ‘correctness’ in English grammar and usage is still with us; there are certainly displays of it within the community in letters to newspapers and authors (such as Truss 2004; Watson 2003; 2004) who set themselves up to be what Cameron 1995 calls ‘verbal hygienists’. Verbal hygienists include language conservation groups, movements for plain language use and spelling reform, those who take elocution lessons and ‘communication’ courses, editing language to a house-style, guidelines for nondiscriminatory language, the mocking of accents, and sanctions against swearing (Cameron 1995: 9). Most of these have aims similar to those of the eighteenth century prescriptivists: 156 The Western Classical Tradition in Linguistics to clean up and ‘correct’ the language, and to teach youngsters what is deemed correct. There is even a moral dimension: If you allow standards to slip to the stage where good English is no better than bad English, where people turn up filthy at school [...] all these things tend to cause people to have no standards at all, and once you lose standards then there’s no imperative to stay out of crime. (Norman Tebbit MP, on BBC Radio 4, 1985; quoted ibid. 94) Bad English is correlated with slovenly unkempt appearance and criminal behaviour. 

If children are drilled in rule-governed grammar, the view held by people such as Tebbit is that they will learn discipline and imbibe a respect for order, tradition, authority, hierarchy, and rules (ibid. 95, 112). There is absolutely no experimental foundation for the view that learning a language or any other rule-governed intellectual behaviour (such as logic or mathematics) constrains subsequent delinquent behaviour. And the everlasting problem of deciding what counts as ‘correct’ grammar persists. Today, as in all earlier times, the evaluation of good and bad language rests in the eye, ear, and brain of the individual or among the social group to which that individual belongs. We may be pleased that the move to non-discriminatory language in the late twentieth century has been remarkably successful; but most attempts to manipulate language usage fail. Language is a social practice and changes will only be made when a particular practice becomes a convention. Roughly speaking, a convention (Lewis 1969: 78) is a regularity of behaviour to which, in a given situation, almost everyone within a population conforms and expects almost everyone else to conform. Moreover, almost everyone prefers this state of affairs to an alternative. This is not to say that the convention is immutable: if people cease to conform to a particular regularity and prefer to cease to conform to it, it will cease to remain a convention; and if they gradually adopt another regularity in behaviour, this will become a convention when almost everyone in the population conforms to it and almost everyone prefers this state of affairs to the alternative. 

Even when language conventions such as forms of address are imposed following a revolution, they do not succeed in being maintained unless they are willingly accepted within a community – as French, Russian, and Chinese revolutionary history has shown us. When language is being learned there must be prescriptions about what is ‘correct’. ‘Standard English’ is the written variety that is promoted in schools and academia, that is used in most non-fiction books, law courts and government institutions. Students are expected to use it in essays; and ESL instructors teach it to foreign learners of English. Writers are supposed to acquire the standard rules and those who do not are in danger of being regarded as recalcitrant, lazy, and incompetent; they are said to have poor command of grammar. Standard English should properly be referred to as the ‘standard dialect’ or ‘standard variety’. There is no standard spoken-English: people reading speeches in Parliament or scripts in media broadcasts read Standard English in the written text, but their spoken delivery uses many accents. Once upon a time the most favoured accent was spoken at Court; then it was the speech of the well-educated upper class, so-called ‘received pronunciation’, the model for ‘BBC English’. Outside Britain there were different accepted spoken varieties such as General American. Although there are regional variations in written Standard English, regional variation in spoken English is extensive and far more obvious. Prescriptivism from the early middle ages on 157 There is a perception that words need to appear in a dictionary before they count as words of the language; hence the power of dictionaries like the Oxford English Dictionary and Webster's Third New International Dictionary (Webster 2000 – the first edition of which was Webster 1828). 

When the project for the publication of a New English Dictionary was proposed in 1857, the aim was to create a dictionary that, based on ‘all English books’, was exhaustive and systematically traced the historical development of the senses and uses of every word, ascertaining its etymology within English and cognates in other languages (Philological Society 1859). This Dictionary shall record, under certain limitations, the existence of every word in the language for which sufficient authority, whether printed or oral, can be adduced, shall investigate its history and derivation, and shall determine, as far as possible, fully and precisely, its several meanings and its appropriate usage, illustrated by quotations. (Philological Society 1860: 3) The OED supplies a written standard for words (and many other dictionaries are afforded a similar status within the community) but there is no comparable institution for English grammar. Nonetheless, many people in the community believe, as their forebears have believed, that lexicographers and grammarians ought to fix the language. One slightly surprising hangover from earlier times is that a study of Latin is supposed to be an aid to the command of good English. The teaching of Latin in schools benefits the English both of those pupils who study it, and those who do not. Standard English has been formed through the centuries by its contact with Latin; and without some knowledge of Latin an Englishman will always remain, to an extent, a stranger to his own culture. [… The teaching of Latin] ensures a supply of English teachers whose grasp of Latin will make their command of English and its grammar firmer and more explicit. (Marenbon 1987: 38) This is pure prejudice. What Marenbon says about the relation of Latin to English applies almost as well to French and somewhat less well to Greek. Marenbon’s error probably derives from the facts that * the amount of Latin most people learn is limited in scope; * Latin is no longer anyone’s first language, and definitive rules can be stated for it without there being a large number of speakers of the language who blatantly violate the proposed rules; * starting with Donatus, history has given us Latin grammars that are succinct. These conditions make Latin appear to be a model of a neat rule-governed language that gives people confidence in their command over its grammar. But such conditions simply will not transfer to the everyday complexities and flexibility of a living language like English. Prescriptive linguistics from the middle ages till modern times The period from 550 to 1100, the early middle ages, did little to advance the Western Classical Tradition in linguistics. Throughout the period the pedagogical grammars of Donatus and Priscian were explicated and developed. From the time of Charlemagne the 158 The Western Classical Tradition in Linguistics teaching of Latin had to be adapted to the needs of speakers of other languages. The result was a model for language teaching that strait-jacketed school grammars of Latin and then grammars of vernacular languages for more than a thousand years. The primers of the early middle ages identify the eight parts of speech in the traditional sequence of nomen, pronomen, verbum, adverbium, participium, coniunctio, praepositio, interiectio, defining each and listing its accidents,28 each property being discussed in turn. During the earliest period the heaviest reliance was on Donatus; but after the carolingian renaissance this extended to Eutyches, Phocas, and Priscian. It was during the early middle ages that Latin teaching was Christianized and principally aimed at the advancement of religion. Throughout that period those being educated in Latin would typically have a passive knowledge of the language from church services, liturgy, hymns, and the Bible – from which a familiar stock of examples were harvested. The Bible was revered for deriving directly from God, which gave the Vulgate authority over everything. The Latin of St Jerome and the other Church fathers was divinely inspired and ipso facto superior to the Latin of any pagan writer. You might think that the difference between classical Latin and Vulgate Latin would have sparked off some interest in historical linguistics, but it didn’t; the early middle ages were not an era of enquiry. To supplement the skeletal grammars of Donatus, which were too underspecified for non-Latin speakers, reference grammars expanding on Donatus and containing lists of paradigms began to be circulated. Many of these are little more than fair copies of teachers’ notes; they were typically less systematic and complete than primers and often limited to noun and/or verb paradigms. Although orthography got some attention, phonology and syntax were virtually ignored. The focus on word-and-paradigm grammar stayed within the pedagogical arm of the Western Classical Tradition until the twentieth century. Priscian’s parsing grammar, the Partitiones, was not widely known until around 800, though it became a favoured teaching model in the post-carolingian period. It was used as a means of instilling the parts of speech and expanding vocabulary and what was then called etymologia – lexical relations (or lexicology); the focus was on derivational rather than inflectional morphology. The study of what we now call morphology was restricted to rote-learned paradigms and no systematic account of Latin accidence was developed. Parsing grammars offer scope for novelty as well as for religious and moral instruction through the text chosen for parsing; they are at the same time easier to update than the primers based on Donatus. The carolingians returned to the pedagogical methods of antiquity in which grammatical study was based on literary texts; but now it was principally the Holy Scriptures and commentaries on them. In the late middle ages this gave way to dialectic, though only at higher levels of study. From the late eighth century, the insular teachers addressing speakers of Germanic or Celtic languages were compiling form-based grammars. Eighth and ninth century Germany, in the Frankish Empire, was Christianized by missionaries from England and Ireland; it was 28. The references here and subsequently to accidentia, accidentaliter, accidents, accidental mode, etc. have to do with secondary grammatical categories typically expressed by inflections. Prescriptivism from the early middle ages on 159 they and their students who preserved insular grammars. A glimpse into the way in which manuscripts were spread is the following: [T]he Ars Tatuini commenced its Continental transmission at a Frankish centre, probably the palace-scriptorium, traveling outward in the company of two different groups of texts, one including several Classical works – Priscian’s Institutio de nomine, Servius’ De finalibus, and excerpts from Charisius – and the other consisting of the early medieval grammars of Augustine, Julian of Toledo and Paulus Diaconus. (Law 1982: 100) Although Boniface was conscious of his obligations to English speaking students, it was another 300 years until Ælfric used English as a medium of instruction. But Ælfric did not go so far as to write a grammar of Anglo-Saxon despite the fact that people already wrote the language and presumably must have been instructed how to do so. Anglo-Saxon used as many as four letters not in the Latin alphabet: wynn [w], thorn [?, ð], eth /ð/, and miniscule yogh , which in middle English become [?, ?, j, ?]. Wynn and thorn are runic in origin and eth and yogh come from Irish. So some training in orthography must have taken place, but there is no record of grammar being taught. By the ninth century there was already re-awakened interest in the rationale behind the parts of speech. Parsing grammars were updated and new topics such as suppositum, appositum, regimen, and constructio introduced. As we shall see in Chapter 8, this movement flourished during the late middle ages in the era of speculative grammars which re-established the theory of grammar in the Western Classical Tradition. We skipped from the middle ages to post-renaissance England, where the Western Classical Tradition is manifest in English grammars that were closely modelled on traditional Latin grammars. There was a belief in the ‘genius’ of English which seems to correspond to what today we call ‘intuitions about grammaticality and acceptability’. Although there was constant harping on the purity of the language, there was a heavy reliance on the grammatical rules found in (or assigned to) the classical languages of the Bible, but especially Latin – which was better known than Greek or Hebrew. It was a view nourished by the training that all grammarians had received when learning Latin in school, but also from a notion that Latin is closer than English to the ‘general’ or ‘universal’ grammar common to all languages, and hence an excellent model for ‘correct’ grammar. Many seventeenth and eighteenth century grammarians believed that general grammar must be logically structured and so they looked to logic as the basis for ‘correct’ grammar. Others believed that language reflects nature, and looked to relations in nature to explain grammatical relationships, as the speculative grammarians had done in the late middle ages. Finally, there was the criterion of ‘good usage’, that is, usage by a ‘reputable person’, an educated gentleman,29 who spoke ‘pure’, i.e. not foreign or provincial, English. This encouraged a subjective judgment that the grammarian’s own brand of English was the correct one. It was the perfect environment for the grammatical bigot. Nineteenth century and early twentieth century linguistics castigated as unscientific the practices of eighteenth century prescriptivism which earned so-called ‘traditional grammar’ a bad name. The very 29. There was a similar criterion on the results of scientific experiments being validated by gentlemen. See Shapin 1994. 160 The Western Classical Tradition in Linguistics negative connotation of the word prescriptivism among linguists in the twentieth century is the reason that Cameron 1995 referred instead to ‘verbal hygiene’, which, despite her disclaimer, is just a euphemism for prescriptivism (Allan and Burridge 2006, Ch.5). As I hope has been made clear in this chapter, there is nothing wrong with prescriptivism in pedagogy; it is a legitimate arm of the Western Classical Tradition in linguistics. Only when prescription aims to clean up the language by subjugating the description of actual linguistic practice to an imposed ‘corrected’ linguistic practice should it be reviled like any other form of censorship.