Of fox-sized mice and a thousand men: Hyperbole in Old English

Claudia Claridge
University of Duisburg-Essen

Abstract

1. Introduction

Did the Anglo-Saxons exaggerate? One might be tempted to find this question superfluous, but given the statements about litotes, or the opposite of hyperbole, found in the literature on Old English writings it may be justified. Hollander (1938) identified litotes as the most Germanic of all rhetoric figures, and according to Campbell (1962: 18f) “the all-pervading litotes is doubtless an ancient Germanic feature”. Scholars discussing the epic (style) in Old English, a style one might think prone to exaggeration, give hyperbole short shrift: Campbell (1962: 19) mentions it only briefly and does not give it much weight, while Harris (1988) denies the world of Beowulf any hyperbolic quality of note. Such statements would seem to indicate that there is very little exaggeration to be found in Old English texts.

However, though the stereotype of modern British English speakers also leans towards a strong tendency for understatement, nevertheless they do exaggerate (Claridge 2010). Exaggeration also seems to be a general human tendency, so that the overt prominence given to one figure does not exclude the other. Anglo-Saxons certainly did not say much about hyperbole. Bede, for example, lists hyperbole among 13 tropes in his Schematis et tropis sacrae scripturae liber, but gives it the shortest treatment of them all. The figure also hardly appears in Knappe’s (1994) comprehensive work on classical rhetoric in Anglo-Saxon England.

Given the state of affairs in research and especially the fact that no study so far has been devoted to the use of hyperbole in the English Middle Ages (in contrast to the German and French situation, cf. Gumbrecht 1972, Oßberger 1973), the present study takes a first exploratory look at the use of exaggeration in Old English writings. After providing a definition of hyperbole (section 2), I will deal with methodological challenges of investigating hyperbole in the type of data specific to Old English (section 3), highlighting in particular the problems of clearly identifying individual uses as hyperbole proper (section 4). A corpus study of selected elements will then be presented in section 5, producing tentative but hopefully indicative frequencies of Old English hyperbole. This and the following more qualitative case studies of selected works (section 6) makes possible a first survey of forms and functions of hyperbole in Old English (henceforth OE).

2. Defining hyperbole

The definition to be used is the one presented in Claridge (2010: chap. 2), which will be succinctly summarized here. In a nutshell, a hyperbolic expression carries magnified, intensified content that exceeds the (credible) limits of fact in a given context. Put more precisely: at the basis of the definition lies the contrast between the hyperbolic expression and an assumed “literal” expression, which would represent the content and context more factually or faithfully. It is this contrast that triggers a transferred interpretation. The contrast needs to fulfil the following conditions:

1. the hyperbolic expression expresses more of essentially the same type of content than the literal one (magnification, realised partly by linguistic gradability), e.g. hyperbolic for ages vs. literal for an hour (content: time),
2. the difference is sufficiently large to stand out/be noticed easily, e.g. try something (literally) five times vs. (mildly exaggerated) eight times vs. (clearly hyperbolic) 50 times,
3. the contrast is nevertheless still reconcilable, i.e. speaker and hearer can clearly see a connection between hyperbolic statement and literal statement/context, e.g. know a lot about sth. vs. know all about something. This normally does not present a problem in straightforward hyperboles (called “basic” hyperbole in Claridge 2010) such as the example in the previous sentence. In composite hyperboles, i.e. those involving another figure such as metaphor or metonymy, reconcilability can pose difficulties, however (e.g. a monster car: hyperbole for size or power or ugliness – or no hyperbole at all).

The transferred interpretation arrived at this way takes up a semantic middle position between the literal and the hyperbolic expression, that is the hyperbolic statement is not completely corrected downwards. The semantic “surplus”, so to speak, expresses a part of the function of hyperbole, the expression of intensity and/or the emotional attitude / evaluation of the speaker towards the content. Describing something large as gigantic expresses that the speaker is impressed by it and its size, either positively or negatively. The speaker is committed to the function of his or her hyperbole, but not to the literal meaning of the hyperbolic item(s). This also means that the speaker is (at least latently) aware of the contrast, and that s/he will always be able to admit that an exaggeration has been used. There is always some kind of intentionality involved in the use of hyperbole.

The original definition works not with a simple concept of context, but with the construct of the “objective observer” and the fact that everything is “seen by” the language users as something or other. This raises a potential problem: we are of course rather far removed from Anglo-Saxon speakers and it is certainly more difficult (to try) to take their perspective. Talking about allegory and some other figures, Stanley (1955: 425) hinted at the possibility that “in some cases it is possible that what seems figure to us was fact to the Anglo-Saxons”. This problem is aggravated by certain types of text; I will come back to this problem below.

Instances of hyperbole can be more or less novel, creative or conventional. The latter leads to another problem concerning OE, namely the fact that we probably do not have enough data available to decide upon the status of a given hyperbolic item in this respect.

3. Data and methodology

Old English data is basically finite; everything that has come down to us is collected in the Dictionary of Old English Corpus (DOEC). This corpus contains poetry, prose, interlinear glosses, glossaries, runic inscriptions and inscriptions in the Latin alphabet, which altogether comes to 3,033,142 OE words (plus 758,503 foreign words). Of those only the first two categories, OE poetry (DOEC: A) and prose texts (DOEC: B), were considered relevant and will be used below. These two sections come to 2,306,261 OE words or three quarters of all attested OE. The prose and poetry section were selected because they represent self-contained text, often of a sufficient length, with some degree of topic development and providing at least some necessary context.

The OE texts will be made use of in two ways: (i) as a corpus to be investigated with corpuslinguistic means, and (ii) as a smaller selection of  individual texts to be read through and scanned for potential occurrences of exaggeration. The reason for this two-fold approach lies in the fact that it allows both for comprehensiveness and for novel discoveries.

Approach (i) means that one has to predetermine search words or phrases for a keyword-in-context search (carried out with Wordsmith); in other words it is necessary to know potential hyperbolic items beforehand. As there is no previous research into OE exaggeration, one can only use research findings on modern English hyperbole in order to decide on search items. Three studies have come up with common semantic fields and items used in modern exaggerations, namely McCarthy and Carter (2004), Cano Mora (2009) and Claridge (2010). McCarthy and Carter (2004) apparently worked with a pre-defined list of semantic fields, which they applied to their data. The four fields of number and quantity, spatial extent, time and degree of intensity thus need not be taken as comprehensive. Cano Mora’s (2009) fields, in contrast, seem to have arisen from the data, but in her case the definition for identifying hyperboles remains somewhat unclear. Her fields are evaluation (which includes positive, negative, and impact/singularity), quantity/purity, quantity/measure (which includes time, space, and number), and magnitude. The semantic domains provided in Claridge (2010: 75), which were extracted from the data, comprise the eight domains value, activity/event, time, quantity, degree, human state, physical property, and dimension. They are based on items used and concepts represented or referred to. It is obvious that there are clear areas of overlap between the three studies: quantity and time occur in all three, while other terms cover the same ground: spatial extent, space and dimension; degree of intensity, degree, and magnitude. Quantity and its related or sub-field number as well as space/dimension were therefore chosen as promising search avenues into the OE data. In order to find search items, the Thesaurus of Old English (2000) was consulted. For quantity, words from the field “an immense quantity, immensity, immense number” (parts of category 03.03.04.01.03, p. 164) were selected and searched for in all possible variants and forms: un(ge)rim ‘countless (number), untold’ , nan rim ‘countless’, and ungeendod ‘unending, endless, infinite, boundless’. [1] þusend was included as a numerical quantity word, as it is apparently a long-standing hyperbolic term (Claridge 2010: 188–193). Dimension words were taken from the subfield ‘huge, immense, enormous’ (part of 03.03.04.01.01 ‘greatness, bigness, size’, p. 162f) and embraced: ormæte ‘boundless, huge, excessive, intense’, ormætlic ‘excessive, tremendous’, ungefōg ‘immoderate, excessive’, ungefōglic ‘excessive’, unmæte ‘excessive, immense, great, vast’, unmætlic ‘enormous’, ungemet ‘huge; immeasurably’, uþmæte ‘huge’. As is obvious from the translations provided, quantity, dimension and degree meanings blend into each other, so that it may not be possible to keep the chosen semantic domains neatly apart. Words denoting ‘death’ have also been shown to play a role in exaggeration, for example in Claridge (2010: 197–207) and Cano Mora (2009), where the concept figures in her ‘negative’ evaluation category. Therefore, the form deaþ and its variants were included in the search. Finally, as comparison plays a small role in modern hyperbole (Claridge 2010: 64–66), one OE possibility for expressing comparisons, swilce ‘as, in like manner, resembling, as if’, was also taken into account. The following lexical/semantic and formal categories will thus be made use of below: the lexical items death and thousand, the semantic field of quantity and dimension, and the rhetorical device of comparisons (cf. section 5, Table 1).

Approach (ii), reading through corpus texts (already used in Claridge 2010 for modern material), makes it possible to find unexpected forms of exaggeration. The texts selected for this are the following:

1. Beowulf, Battle of Maldon, Battle of Brunanburh
2. Apollonius of Tyre, Letter of Alexander to Aristotle,
3. Wulfstan’s Sermo Lupi ad Anglos, selections from Aelfric’s Catholic Homilies. [2]

The texts in (a) represent the epic poetic tradition, with the latter two having links also to historical writing (Brunanburh appears in the Anglo-Saxon Chronicle, entry for 937). The setting is the Germanic world, in the latter two cases England itself. The two texts in (b) are prose narratives (with descriptive elements), both set in worlds far removed from the Anglo-Saxon experience, namely classical antiquity, the Mediterranean and the East. They are similar to Beowulf in so far as they involve fictionality, but how Anglo-Saxons regarded the relationship between fiction and historicity in these particular texts must ultimately remain a guess. The sermons in c) present a didactic genre, containing elements of teaching and persuasion.

OE sources may generally not be unproblematic for investigating hyperbole. First of all, they are all written sources. Presumably (although without proof so far), exaggeration is more natural and thus more common in spoken language. Kreuz et al (1996) have nevertheless found hyperbole to be the second most frequent figure (after metaphor) in modern short stories, a written genre. Written sources can thus be useful, even though they might lead one to more established, bleached examples – or to consciously rhetorical uses, which given the statements in the introduction is not very likely for OE exaggeration. Secondly, there is the problem of register: quite a few of the attested OE registers may not lend themselves easily to hyperbole. For example, legal and historical writing may be too sober or too matter-of-fact for exaggeration, and religious writing may be too serious or earnest in outlook for it. Thirdly, religious writing shares with heroic (e.g. Beowulf)  and “fantastic” literature (e.g. Wonders of the East) the fact that their subject matter may make it hard to distinguish between what counts as real, as an accepted fact or as hyperbole within their textual worlds.  Finally, there is the problem of translations, which raises the question of how “indigenous” to OE are the hyperboles found such texts. Overall, there remains the problem of context. As seen in § 2, hyperbole is essentially defined with reference to the (extralinguistic) context, which makes knowledge about it, even beyond the immediate textual world, indispensable. This is not always given or easy in the context of OE literature, as I will show below.

4. Identifying hyperbole

In this section I want to point out some problems one can encounter when trying to unambiguously identify instances of exaggeration in OE texts. Two of them, historical context and religion/the supernatural, have at least partly to do with the specific Anglo-Saxon context, while the third, collocations and hyperbolic potential, is of a more general nature.

Starting with historical context or common world knowledge, we find examples such as the following from the Anglo-Saxon Chronicle, indicating vaguely that “many thousands” (1) or, more precisely, 840 people (2) were killed.

Does the vagueness in (1) indicate hazy overstatement while the precise numbers in (2) speak for realistic figures? Or does (1) simply mean that given the circumstances no precise figures were available for the writer and he had to make a guess? Entries in the Anglo-Saxon Chronicle are more often vague than precise with regard to numbers, and it is probably more in line with the general style and spirit of the annals to assume informed guesswork than widespread intentional hyperbole. Whatever the interpretation, it is well known that numbers in old historical sources are often not to be taken at face value. For the modern researcher it would be helpful to know the size of contemporary armies to put the figures into perspective. Armies in the ninth century are said to have numbered in the thousands (Stenton 1971: 243, fn.1), but as the modern estimates are at least partly based on sources like the Anglo-Saxon Chronicle the argument easily becomes circular. If we have armies of several thousands and “many thousand” were killed, we have either a disastrous battle or a case of overstatement – but which of the two is not solvable by linguistic means. An additional linguistic observation might help, however: the phrase feala þusend recurs several times in the Chronicle, which might speak for it being a formula, a fixed phrase with perhaps no precise semantics but the function of indicating something like “serious or important battle”.

The numerical information in (3) at first glance seems less problematic, as it should be possible to establish the œuvre of Augustine of Hippo unambiguously.

According to modern counting, Augustine wrote about 100 books. But did one count in the Middle Ages in the same way? What precisely counts as “one” book then – physically or conceptually? The City of God alone, for example, is nominally subdivided into twenty-two “books”. However, even with a generous counting one would probably not reach 1000. [4] The fact that the same figure occurs also elsewhere in OE (e.g. the Exeter list of relics) points to it being a commonplace and potentially a conventional hyperbole.

Like the the army examples above, the following examples make extralinguistic knowledge necessary, which here is the knowledge of the extent of the OE measure unit foot. The same is the case for all contexts involving old measurement terms, such as foot in the following examples or mile in (26) below.

There is apparently no certainty about the OE foot in historical and archaeological research. There are two fairly close estimates, arrived independently in Jarrow (Bettess 1991) and in Yeavering (Hope-Taylor 1977), which put it at 280 or 281 millimetres, whereas Sunley (n.d.) puts the foot at 335 millimetres. Taking the former as the basis the people in (4) would be about 2.50 metres, while the bishop in (5) is about 2.80 metres tall. This certainly exceeds the Anglo-Saxon norm (for comparison: a man of 1,8 metres would then be about 6,4 OE foot), if only slightly. Taking the larger estimate one gets people of about 3 metres and a bishop of 3,35 metres, which is perhaps a better hyperbolic contrast. [5] The textual context here raises a further question. The Letter of Alexander describes a far-away, exotic world, in which quite a few things do not fit assumed Anglo-Saxon norms. Is the deviance thus hyperbole with the function of creating a strange (fictional) world, or is it taken as a true rendering of a foreign world, as fact? And how unusual would the fact then be? The trees in (6) with ca. 28 or 33,5 metres height are either actually possible also in a European context (roughly the size of a poplar or a plane) or only very slightly higher, but the context seems to imply that the epithet wunderlicre heanisse also applies to them and thus marks them as unusually high. It is hard to clearly determine whether these passages involve hyperbole on the basis of our incomplete contextual knowledge; the one linguistic fact that may tilt it towards hyperbolic use (but also any other symbolic use, for that matter) is the use of round numbers in (5) and (6) (cf. e.g. McCarthy & Carter 2004). [6]

The second problem area mentioned above is religion. Religious belief as such deals with the unusual, the supernatural, the highest matters, all-encompassing concepts etc. – is it possible at all to exaggerate in this field? Take (7), for example, where the devil is compared to a dragon. While this is a potentially graphic visualisation (depending on the Anglo-Saxons’ ideas of dragons, of course), it is certainly not hyperbolic, rather the opposite. With the devil being the most extreme possible evil, he can hardly be further exaggerated.

The same goes for the heavenly side, which is featured in (8). The involvement of God will have made it clear in a Christian context that one will never have seen or heard anything of the like before – in the earthly contexts one is familiar with. Thus, while the form (“never VERBED”, potentially repeated with various (similar) verbs) is commonly found in (modern) hyperbolic instantiations, it cannot be taken as hyperbolic here. In so far as homiletic writing prominently deals with the Biblical and other-worldly personages, little hyperbole can probably be expected in this area. Only where religious texts become more hearer-centred in their content might there be more scope for exaggeration (which is one reason for the choice of sermons here). Similarly, problems arise in texts that deal with mystical and fantastical subjects, whether religion-related or not. If the neck of the phoenix is described like polished gold (his sweora swylce smete gold, Hom U 17,2), this is not “testable” against real context, as this type of bird simply does not exist.

The last aspect to be discussed here concerns the common use of words as attested by their collocability. Collocations can be crucial in determining the usual meaning of a word; only if one knows this can one determine whether a given usage might deviate from this default meaning, perhaps towards hyperbole. This might help decide, for example, whether the boars in (9) are just large within a reasonable range (given contextual modulation of meaning), or whether they are described as large to an unusual degree or hyperbolically. Looking at the collocations of the unmæte-family is not very helpful, however. Apart from army, swamp, stone, wild beasts, folk, and forest, there are no other concrete collocates that could help delimit the default type of size the word indicates. [7] None of those leads one to necessarily conclude a truly large size. This is in line with the context of (10) where the beam of great size is actually equal in height to a man. Thus, the only aspects that might speak for the boars (9) being unusually large is the immediate textual context where, for example, lions the size of bulls and bats as large as pigeons are mentioned, and that unmæte in its intensity sense in other places denotes something extreme (for example the torments in hell).

Ormæte (11) is a slightly different matter, as it has more collocates of a concrete nature and, crucially, collocates with words like giant (referring, for example, to Goliath, Hercules) or elephant, and appears in the description of the wealth of Job, “the greatest man in the East” (cf. also AV Job 1,3). Thus, although ormæte also collocates with stemn “voice” several times, the simile involving a huge trumpet can probably be taken as hyperbolic.

In conclusion, one can say that the identification of OE hyperbole is more difficult than that of modern instances because through the lack of common world knowledge and through non-native linguistic competence it is much harder or almost impossible to rely on intuitions. Extensive historical and/or linguistic research will be necessary for conclusive judgments on many examples.

5. Corpus study

For the corpus study, keyword-in-context searches were conducted using truncation and wildcards in order to find as many instances as possible. This procedure finds not only deaþ, but also, for example, deaþlicness and deaþscyldig, thus not only taking care of inflected forms but also potentially extending the results to word families. Table 1 shows the outcome of these searches, giving the total occurrences found for each word family or semantic field in the selected corpus and the instances which could be understood hyperbolically. Only fairly unambiguous hyperboles were included.

It is apparent that the amount of hyperbole is small indeed. But how does it compare to modern data?

• If one compares the OE result for the death-group simply with all hyperbolic occurrences of the phrase to death (as in laugh oneself to death) in the spoken BNC (cf. Claridge 2010: 201), one finds the OE result dwarfed by the modern occurrence of 0.64 per 100,000 words.
• For thousand(s), Claridge (2010: 60) found that 0.25% of singular instances and 0.8% of all plural instances in the spoken BNC [9] are hyperbolic. In the OE data, the percentage is actually higher with 1.6% hyperboles. This may have to do with the fact that number hyperbole of this type has a long and vital history, also having notable uses in the Bible (cf. ibid. p.188–193).
• Comparison, in contrast, is clearly less prominent in OE than in the modern data: in the subcorpus of the BNC used in Claridge (2010) hyperbolic comparisons occur with a rate of 4.1 instances per 100,000 words, whereas they come to only 0.2 per 100,000 in the DOEC. The modern comparative structures included in the count are more varied (like, than, as, tantamount to etc.), however, while here only swilce was investigated.
• The semantic domains quantity and dimension taken as a whole (i.e. with many different realisations) also come out as much more prominent in the study of the BNC subcorpus in Claridge (2010), with 13.0 and 6.25 instances per 100,000 words respectively.

Needless to say, the OE-PDE comparison is based on an unequal footing. For three of the above cases (death, comparison, quantity/dimension) modern speech is compared to OE writing, while in the remaining case (thousand) the whole multi-genre/multi-register BNC is compared to the restricted range of OE text production that has come down to us. A possible approach for a future study might be to use a modern corpus that is more similar to the OE material (e.g. sermons, certain types of history writing) in order to increase comparability. The inclusion of more and more varied features is another option; however, thousand and comparison are already features with an affinity towards written, literary uses. Thus, the differences shown above are nevertheless striking enough and seem to corroborate the Anglo-Saxon disfavouring of exaggeration mentioned in previous studies on OE literature.

Looking at the most promising result first, i.e. þusend, one finds this use most prominent in religious sources. Partly this may be due to foreign models, most clearly in the case of biblical or similar foreign-language originals, as is the case in examples (12) and (14), which contain faithful translations of their sources.

Thousand in (12) and (13) emphasizes the safety of the believer, whom nothing worldly can truly hurt anymore; in other words, it highlights the strength faith provides. In (13) this is made more graphic by the intensification from 1,000 to 10,000.  In (14) it is the great sinfulness of man that is stressed, which in turn puts in focus the mercy of God towards humankind. That is, in each case the use of thousand helps to make a theological point more clear. A hyperbolic use is often close to (or even identical with) a more general symbolic use (cf. Dobrovol’skij and Piirainen 2005: 286–9). Ælfric comments on this metalinguistically in one of his sermons in the context of expounding the Biblical story of the feeding of the five thousand:

Similar uses to the above are found in other texts. Both (16) and (17) also call attention to the sinfulness of man, in these cases in particular people who do not (yet) behave as a good Christian should.

In both cases, thousand is further highlighted and made truly hyperbolic by the contrasts in the context: a thousand sins in one day and no penitence at all, and one belly and food for a thousand men. The purpose in these sermons is clearly persuasive and instructive. By highlighting the unwanted behaviour to the extreme, the aim is to make such conduct repugnant to the listeners, and thus ultimately to make them change their own behaviour where necessary.

The last example with thousand to be presented here, (18), has an emotive function familiar from modern hyperbole.

In (18) the Soul is speaking to the Body, expressing its negative assessment of the situation of being trapped within. A time hyperbole is employed to mark the subjective pain. Both are fairly common in modern usage, where in particular negative evaluations are prevalent (Claridge 2010: 76, 81); this also fits the state of affairs in (14), (16) and (17) above. Me þuhte in (18) functions as a hedge, which is something also found in modern hyperbole without necessarily diminishing the hyperbole (Claridge 2010: 102–105); it simply increases the degree of subjectivity.

Comparisons seem to be quite common in OE, but mostly they are obviously not of the hyperbolic kind. The hyperbolic nature of (19) comes about through combination of teh phrase eall wæs besæt with a hedgehog simile, because then the “innumerable spines” (cf. OED definition of hedgehog), i.e. the amount and closeness of the spines, are foregrounded. It could almost be argued that the visualisation achieved adds a humorous dimension, but this was probably not intended by Ælfric.

The contrast between stone and sand in (20) highlights the ease of the man’s action, i.e. the comparison works like an intensifier. The swilce-clause in (21) emphasizes the greatness of the fear, thus being reminiscent of the modern phrase frightened to death. For illustration of other possibilities of OE comparisons (not included in Table 1), see gelic in (22) and swa…swa in (23).

Both are hyperbolic in nature, so that there is further scope for exaggerated comparisons in OE.

Quantity and dimension searches did not yield great surprises. The two examples from the Letter of Alexander work with the concept of uncountability, i.e. the quantity is so great that it is impossible to do any counting.

In most cases, this automatically produces hyperbole, as counting is not truly impossible. This use is still around today in conventionalized and weakened form. In (25), the uncountable-trope is supported by a simile, pointing to a situation where counting would really be impossible. Similar to non-countability, there is also incredibility, that is, something is so great/much that it is (almost) impossible to believe it; this is used to introduce the following passage about the city of Babylon from Orosius’ history:

Here, the huge dimensions of the fortifications (i.e, the walls and the moat) are described. The circumference of circa 70 miles corresponds to Orosius’ (ultimately Herodotus’) 480 stadia, which is actually 60 Roman miles (Bately 1980: 235). The figure of 70 used by the OE translator may have been based on a calculation with a smaller mile (ibid.). The other measurements as well as the “unbelievable”-phrase were probably also faithfully taken over. What the translator seems to have added to the original text, however, are the two absolute superlatives, mæsta dic and  ungefoglecesta stream (Claridge 2010: 62), which may serve to further mark the extraordinariness of the city. Modern excavations have shown Herodotus’ measurements to be far too large , so that we have either a case of hyperbole or of misinformation. [11] What the translator thought of the content, whether he believed it and whether the added superlatives simply emphasize the perceived truth or are rather thrown in as empty, perhaps distancing, formulae is impossible to know. This example, as well as (24)–(25), highlights the problem of translated literature: some things may look like hyperbole, may not fit reality, but nevertheless the intention of the OE writers must remain unclear.

Similarly to “countless” and “incredible”, claiming infiniteness as in (27) with ungeendod can also signal hyperbole. The hedge fornean ‘almost’, like me þuhte  in (18) above, does not destroy the hyperbolic effect completely. Moreover, the preceding intensified phrase supports a high-level interpretation. The hyperbolic scenery here underlines the supernatural context involving an angel.

Ælfric here relates a passage from Bede’s Ecclesiastical History, with very similar phrasing (in particular the item infinitae).

As seen in Table 1, death is the least common hyperbolic term in the corpus. Of course, it would be necessary to also check the adjectival and verbal forms in this field in order to confirm or disprove the rarity of hyperbolic death. But it is nevertheless tempting to speculate that a society or culture that knows death at closer range than ours may perhaps not be so tempted to use the concept loosely. The two examples found are in fact originally foreign sources: one from Orosius (28), the other from Gregory’s Dialogues.

What we find in (28), in the original as well as the OE version, is a form of “rather die than do x”, indicating (more or less hyperbolically) that the alternative, here making peace according to Roman conditions, is worse than death. Phrasing a denial this way shows both outrage (at the impossible offer) and determination not to yield at any cost. In this context, words indicating ‘death’ are of course also a factually logical choice (beyond the rhetoric effect), because refusing the peace offer meant fighting on and thus death for at least some Carthaginians.

As the corpus approach did not yield many results, and in particular no surprising ones, I now turn to the second approach of finding hyperbole in OE.

6. Case studies

The purpose of reading through complete texts was to find unexpected hyperboles, but also to get a better feeling for the extent of use and for the functions of hyperbole in context. However, nothing unexpected or truly unusual was found by this method.

I will briefly characterise four of the selected works, and concentrate especially on forms not discussed so far.

Beowulf

Perhaps surprisingly, there is not much hyperbole in Beowulf – at least not of the type that one can nail down to specific linguistic forms. One might of course call (some of) the plot hyperbolic in nature, but this is not the cumulative result of many individual hyperbolic items.

The instantiations that are found are also not very spectacular in form or conceptualisation. There are, for example, a number of superlatives, partly of the absolute, unmodified kind, such as husa selest ‘the finest of halls’ and irena cyst ‘the most select of weapons’, partly of the modified kind as in foremærost foldbuendum receda under roderum ‘the most eminently celebrated among earth’s inhabitants’ and moncynnes mægenes strengest on þæm dæge þysses lifes ‘the strongest of mankind at that time in this life’ (cf. also Claridge 2010: 247). They are listed here in decreasing strength, with the last one with its curious restriction being least hyperbolic. Superlatives are fairly unobtrusive as hyperbole goes; they will not strike the reader as excessively hyperbolic, in particular if they are fairly rare. The following example, occurring in the context of Beowulf’s rendering of the swimming contest, uses several combined means to produce an extreme-case statement: negation, comparisons, universal adverbial, contrast, and a common formula.

Universal descriptors occur more often, cf. under roderum ‘under the skies’ above or on foldan gefrægen hæbbe ‘have heard on earth’, the latter also together with the formula, as they do in modern English (in the world, on earth). Negation is also more widespread; it occurs in the Battle of Brunanburh as well, as in for example the following: Ne wearð wael mare […]  siððan eastan hider Engle and Seaxe upp becomon, ‘never was there more slaughter … since the Angles and Saxons came from the east’ (Kinsella 2009: ll. 65–70). Items like never, and its positive counterpart ever, can also function hyperbolically in modern English. (29) highlights the prowess of Beowulf, as do other passages, like the description of him carrying thirty war-harnesses while swimming (hæfde him on earme ana XXX hildegeatwa), which is an activity and quantity hyperbole combined. The difference is that (29) is spoken by Beowulf himself and is part of a boasting ritual where hyperbole has a natural place, whereas the just quoted element is a third-person description. Moreover, the high quality of weapons is highlighted in, for example, a helmet that neither sword or battle-knife might cut through (helm … þæt hine syðþan no brond ne beado-mecas bitan ne meahton). [12] Apart from the two last quoted examples, most instances found in Beowulf make the impression of being fairly expectable and conventional usages.

The rare hyperbole in Beowulf thus centres on the protagonist himself and on fighting equipment. Beowulf does stand out, but he is not too much elevated linguistically. It is as if the poet wanted to present a largely “normal” world, which is supernaturally threatened from the outside: neither Grendel nor the dragon are described hyperbolically, perhaps because they cannot be linked to the natural world (cf. the context-link of hyperbole). What may also be noteworthy is a potential ironic-subversive touch to the hyperboles used: the highly praised objects are either under fatal attack (the hall Heorot) or fail in moments of danger (the sword Hrunting). The scarcity of hyperbole, in particular vivid, creative types, might also be linked to a further aspect, namely the general lack of visual characteristics of Beowulf that has been noted in the literature.

Letter of Alexander

It is in texts like this one (also the Wonders of the East, for example) that the problem of context presents itself in particular. The world described has “fantastic”, imaginary qualities, but it is an earthly world described – reputedly – by Alexander the Great for his tutor, Aristotle, that is, the participants are two respectable, trustworthy persons. Were the contents fact or fiction for the Anglo-Saxons? If they were fact, constructions we rate as hyperbolic might not have been so for the Anglo-Saxons; if they were fiction, the hyperbole found has a function similar to that found in modern tall tales and magical-realist fiction (Claridge 2010: 253–256), namely that of establishing a certain kind of world that is set apart from other worlds, but nevertheless not completely unusual. It is a world that has bishops, but they can be extremely old, as in (30), thus combining the expectable (the existence of bishops) with the unexpectable (their high age).

Here, this type of world serves to highlight the great achievement of Alexander, i.e. his courage and persistence in adventuring into unknown and potentially dangerous areas – in other words, it can be seen as a subtle means of boasting. Nevertheless, the text does not contain much clearly identifiable hyperbole (cf. the examples discussed in § 4 above, for example), and what is there is unspectacular. There are a few size and quantity hyperboles, the latter mainly of the type “innumerable”, “countless”. While these are probably conventional, (31) may be more imaginative:

Assuming the Anglo-Saxon whale to refer to a large, ocean-living creature, the rivers would have to be fairly large as well. Furthermore, there are a few comparisons, not all of which are unambiguous as hyperboles. (32) could be seen as size-hyperbole, but the mice are also likened to foxes with respect to shape (gelicnisse).

It is therefore not possible to simply see them as (hyperbolic) mega-mice, but also as something qualitatively different.

Apollonius of Tyre

Apollonius shares with the previous text that it deals with a world far removed from the Anglo-Saxon context, but it stays within the circle of the well-known classical world. Scene-setting hyperbole is thus perhaps not necessary, but the little hyperbole found serves the story. As regards the forms and types, there is again nothing unusual from the modern perspective. (33) presents the “incredible” beauty of the girl that triggers the whole plot, and how this attracted people from “every quarter”, using one of the universal descriptors which are also very common in modern hyperbole (Claridge 2010: 51, 52) and figure prominently in “extreme case formulations” (Pomerantz 1986). There are about twenty occurrences in the OE corpus of the word ungelifedlic, several times in combination with to secganne, which may point to a degree of establishment.

This description motivates why Apollonius, Prince of Tyre, is equally attracted to the girl. The reference to literally impossible over all the earth in (34) makes clear the extremely dangerous situation of Apollonius, who is being hunted down.

Similar phrases, such as world-wide, are common in modern English, too, often in conjunction with superlatives. The latter also occur hyperbolically in Apollonius as well as comparisons of various kinds, like the one in (35). Here it is combined with another universal descriptor (“never”).

The overt function of the hyperbole here is praise (as in paying compliments), as it is spoken by a king after he had met Apollonius (the young man) at the bath. But it also marks the return of good fortune to Apollonius after he had been shipwrecked, and thus a turning point in the story.

Sermons, in particular Sermo Lupi ad Anglos

The six sermons by Ælfric, which were consulted, yielded exactly one case of hyperbole:

If pressed, Ælfric would perhaps have admitted that the odd Viking or Pict did not actually believe in the Christian God, so that “all creatures” is an overstatement. The lack of hyperbole concurs with the general scarcity of figurative or colourful language in Ælfric; devices such as metaphor, personification or litotes are used sparingly and, if at all, in a manner that almost literalizes them (Knappe 1994: 392–3). The aim of Ælfric is perhaps rather to instruct than to emotionally arouse.

This is different in Wulfstan, whose aims in the Sermo Lupi ad Anglos are clearly emotive involvement and persuasion. For this he pursues a general maximisation technique, which involves a heavy use of intensifying words (such as swiþe, eall, to), compounds with intensifying first element, a high instance of quantity words (e.g. fela, manige), words indicating extremes and/or universality, and often synonyms in groups reinforcing each other (cf. also Knappe 1994: 396). [13] The instances in (37)–(39) illustrate his techniques: universal descriptions (e.g. dæghwamlice, geond ealle þas ðeode, ælc, ne ænig, gewelhwilcan ende, inne ne ute), intensifiers (ealles to, swiðe), and intensifying doublets/synonyms (oft & gelome, unriht/unlaga) as well as number hyperbole (tyne, twentig in contrast to an, twegen).

The effect is usually not due to any individual usage, but to the combined effect of several items, which amounts to a piling technique. Another notable part of this piling is also seen in Wulfstan’s predilection for excessive listing, as is illustrated in (40), giving the impression of ‘countless’ sins and misdeeds.

The impression of overstatement thus arises out of an overdose of many combined extreme and intense formulations. Matching Wulfstan’s description against what we know of the historical context confirms the hyperbole. Æþelred’s reign was not the best of times and he not the best of kings, but parts of the government and administration were still functioning reasonably, making Wulfstan’s picture of general civic unrest and chaos unlikely. Also, his contemporaries Ælfric and Byrhtferð do not show us a comparable picture of life at the time. Wulfstan thus clearly exaggerated the state of affairs in order to make a lasting emotional impression and to persuade his hearers to a more Christian life.

Wulfstan’s style in the Sermo Lupi certainly stands out, but similar features are not completely absent from other sermons. One example may suffice here:

The writer here also used a piling technique through listing and negation, thereby building up a picture of complete absence. The emphasis is on a completely negative state of emotional affairs.

7. Conclusion

The Anglo-Saxons did exaggerate – at least sometimes and some of them. Ælfric might not have used hyperbole at all, while Wulfstan could indulge in a more or less hyperbolic style. Clearly identifiable linguistic exaggeration is not a hallmark of the epic style (it is not only sparse in Beowulf, but also in the other poems investigated), confirming the statements in the literature. When hyperbole is used in OE texts investigated here, it is not done blatantly or prominently: the preference is for unobtrusive, probably fairly conventional, realisations. The forms found reveal nothing surprising or specifically characteristic of OE: superlative, universal descriptors, negation, comparison, quantity terms, exaggeration within contrasts, and even the piling technique are all found in modern English as well. In spite of the different nature of OE and modern texts, the functions fulfilled by the OE instances are familiar from modern data. OE exaggeration has been seen here to serve graphic emotional expression (e.g. (18)), to praise and to criticize by magnifying either positive (Beowulf) or negative (Wulfstan) aspects,  in the latter case also to further persuasion, and to aid in creating or characterising a foreign world (as in the Letter of Alexander). What sets the OE data apart, however, is the problem of identifying hyperbole with a similar degree of certainty as in modern English.

Notes

[1] Meaning paraphrases taken from Clark Hall.

[2] Selection: Shrove Sunday – Easter Sunday, i.e. X–XV in Thorpe’s edition.

[3] If not otherwise stated the translations are mine. I’d like to thank Matti Kilpiö for checking the translations for adequacy.

[4] A reviewer remarked that Augustine’s Retractiones may be taken to indicate more than a thousand items (“books”) written by Augustine, but I find this hard to verify. The Retractiones certainly mention several hundred works or parts of works (terms used:  opus, liber), but not a thousand. See for example http://www.augustinus.it/latino/ritrattazioni/index2.htm or the copies on google books.

[5] Another aspect that may speak for the people being conceptualised as very tall is their apparently habitual fishing of whales, cf. example (31), § 6 below.

[6] A reviewer raised the question of how one can know what counted as a “rounded” number in OE. Of course, one cannot be sure about that, but the number systems of Old and Modern English are similar enough to at least warrant the assumption. Von Mengden (2010) explains how 10, 100, 1000 serve as bases  in the English numerical system, thus giving these numbers a special status.

[7] The other collocates are abstract (e.g. cruelty) or less amenable to the dimension meaning (e.g. rain).

[8] The two domains were combined here, as the words selected ultimately have both types of meanings, which are not easy to sort apart in every given instance.

[9] The figure for the singulars is based on a random selection of 2,000 from an overall of 5,347 hits.

[10] The hyperbole in the last sentence is already present in the Latin passage. There are Latin sources for several hyperboles of Ælfric’s.

[11] The actual length of the walls is about 18 km or 12 miles.

[12] It has been noted that this description might also point to magical qualities of the helmet, in which case, taking the supernatural world for granted, this form would not be reckoned  a hyperbole.

[13] All intensifiers together make up 49.5 instances per 1,000 words.

Sources

Dictionary of Old English Corpus. 2009 / 1981. Original release 1981 compiled by Angus Cameron, Ashley Crandell Amos, Sharon Butler, and Antonette diPaolo Healey; 2009 release compiled by Antonette diPaolo Healey, Joan Holland, Ian McDougall, and David McDougall, with TEI-P5 conformant-version by Xin Xiang. Toronto: DOE Project.

Bately, Janet, ed. 1980. The Old English Orosius. Early English Text Society S.S.6. London/New York/Toronto: Oxford University Press.

Clemoes, Peter, ed. 1997. Ælfric’s Catholic Homilies. The First Series Text. Early English Text Society No. S.S.17. Oxford: Oxford University Press.

Fulk, Robert D., ed. 2010. The Beowulf Manuscript: Complete Texts and the Fight at Finnsburg. Cambridge, Mass.: Harvard University Press. (Letter of Alexander)

Godden, Malcolm, ed. 1979. Ælfric’s Catholic Homilies. The Second Series Text. Early English Text Society SS.5. London/New York/Toronto: Oxford University Press.

Goolden, P., ed. 1958. The Old English Apollonius of Tyre. London: Oxford University Press.

Kinsella, T.E. 2009. The Battle of Brunanburh. Absecon. http://loki.stockton.edu/~kinsellt/litresources/brun/glossary.html. Link straight to line 65: http://loki.stockton.edu/~kinsellt/litresources/brun/glossary.html#65.

Mitchell, Bruce, ed. 1974. The Battle of Maldon and Other Old English Poems. Transl. by Kevin Cressley-Holland. London: Macmillan.

Plummer, Charles & John Earle, eds. 1892–99. Two of the Saxon Chronicles Parallel. 2 vols. Oxford: Clarendon.  (Parker Chronicle, entry for 937: Battle of Brunanburh)

Thorpe. Benjamin, ed. 1983 (1844). Ælfric. Sermones Catholici, in the Original Anglo-Saxon with an English Version. Vol. I. Ed. Hildesheim etc.: Georg Olms.

Wrenn, C.L., rev. W.F. Bolton, eds. 1973. Beowulf. London: Harrap.

Whitelock, Dorothy, ed. 1976. Sermo Lupi ad Anglos. Rev. Edition.Exeter: University of Exeter Press.

References

Bettess, F. 1991. “The Anglo-Saxon Foot: a computerised assessment.” Medieval Archaeology 35: 44–50.

Campbell, A. 1962. “The Old English epic style”. English and Medieval Studies Presented to J. R. R. Tolkien on the Occasion of His Seventieth Birthday, ed. by Norman Davis & C.L. Wrenn, 13–26. London: Allen & Unwin.

Cano Mora, Laura. 2009. “All or nothing: A semantic analysis of hyperbole.” Revista de Lingüística y Lenguas Aplicadas 4: 25–35.

Claridge, Claudia. 2010. Hyperbole in English. A Corpus-based Study of Exaggeration. Cambridge: CUP.

Clark Hall, J.R. 1960. A Concise Anglo-Saxon Dictionary. 4th ed. Toronto etc.: University of Toronto Press.

Dobrovol’skij, Dmitrij & Elisabeth Piirainen. 2005. Figurative Language: Cross-cultural and cross-linguistic perspectives. Amsterdam etc.: Elsevier.

Gumbrecht, Hans Ulrich. 1972. Funktionswandel und Rezeption. Studien zur Hyperbolik in literarischen Texten des romanischen Mittelalters. München: Fink.

Harris, A. Leslie. 1988. “Litotes and superlative in Beowulf”. English Studies 1: 1–11.

Hollander, Lee M. 1938. “Litotes in Old Norse”. Publications of the Modern Language Association of America 53(1): 1–33.

Hope-Taylor, B.  1977. Yeavering: an Anglo-British centre of early Northumbria. London: HMSO.

Knappe, Gabriele. 1994. Traditionen der klassischen Rhetorik im angelsächsischen England. Heidelberg: Winter.

Kreuz, Roger J., Richard M. Roberts, Brenda K. Johnson & Eugenie Bertus. 1996. “Figurative language occurrence and co-occurrence in contemporary literature”. Empirical Approaches to Literature and Aesthetics, ed. by Roger J. Kreuz & M.S. MacNealy, 83–97. Norwood: Ablex.

McCarthy, Michael & Ronald Carter. 2004. “‘There’s millions of them’: Hyperbole in everyday conversation”. Journal of Pragmatics 36: 149–184.

Oßberger, Josef Franz v. Paula. 1973. Die Verwendung der Hyperbel in der mittelhochdeutschen Epik. Ph.D. thesis, University Erlangen-Nürnberg.

Pomerantz, Anita. 1986. “Extreme case formulations: A way of legitimizing claims”. Human Studies 9: 219–229.

Roberts, Jane & Christian Kay, with Lynne Grundy. 2000. A Thesaurus of Old English. Vol. I: Introduction and Thesaurus. Amsterdam & Atlanta: Rodopi.

Stanley, E. G. 1955. “Old English Poetic Diction and the Interpretation of The Wanderer, The Seafarer and The Penitent's Prayer”. Anglia 73: 413–466.

Stenton, Frank. 1971. Anglo-Saxon England. Oxford & New York: OUP.

Sunley, Harry. n.d. “Romanesque linear measures of England and continental Europe.” March 2012. http://thebaa.org/wp-content/uploads/2014/01/v-Sunley-report.pdf.

Von Mengden, Ferdinand. 2010. Cardinal Numerals: Old English from a Cross-linguistic Perspective. Berlin: Mouton de Gruyter.