Old English Keywords in Context. A Quantitative and Qualitative Corpus Analysis
I will also demonstrate some problems coinciding with the corpus compilation, before offering approaches to the problem of how to gain key words in a diachronic text. Afterwards, I will present the most frequent key words and give a first short interpretation of the results.
All key terms are to be arranged in different semantic fields with respective subcategories and afterwards subjected to a quantitative analysis within these respective fields. Within this analysis I will start with comparing the different fields with each other, then I will compare the subcategories and finally examine the relations within the subcategories. Finally, a qualitative analysis will be performed, once again within the different semantic fields. In this respect, I will also examine the characteristics of pronouns in a separate section.
Table of Contents
2. The corpus
3. Old English key words
3.1 Approaches to gain Old English key words
3.2 Semantic fields
3.3 Quantitative analysis
4. Old English key words in context
In the course of this paper, Old English key words are going to be analysed with regard to their frequency and their characteristics in the respective context. First of all, I will present the compilation of the corpus and argue why the respective sources have been chosen. I will also demonstrate some problems coinciding with the corpus compilation, before offering approaches to the problem of how to gain key words in a diachronic text. Afterwards, I will present the most frequent key words and give a first short interpretation of the results. All key terms are to be arranged in different semantic fields with respective subcategories and afterwards subjected to a quantitative analysis within these respective fields. Within this analysis I will start with comparing the different fields with each other, then I will compare the subcategories and finally examine the relations within the subcategories. Finally, a qualitative analysis will be performed, once again within the different semantic fields. In this respect, I will also examine the characteristics of pronouns in a separate section.
2. The corpus
The corpus to be analyzed is self-compiled, consisting of Old English texts in Modern English translation. The total figure for the tokens is 262.439, derived from 12.912 types. Thus, there is a type-token-ratio of roughly 0,05. Within the compilation, a wide range of texts was included: The heroic poem “Beowulf” with 23.195 tokens, the historic poem “The Battle of Maldon” (2.814 tokens), The Riddles from the Exeter Book (8.218), “The Iunus Manuscript” (38.031) and “Judith” (2791). Furthermore, two chronicles, “The Ecclesiastical History of the English Nation Book I to V” by Bede (115.159) and “The Anglo-Saxon Chronicle” (60.450) were added. The Corpus is completed by a collection of legal texts, “The Anglo-Saxon Dooms” (11.781). Hence, the whole range of literature was included, reaching from fictious to non-fictious, and from biblical to non-biblical texts. Besides, the corpus contains different genres and text forms from poetry to chronicles. The integration of legal texts is justified by the assumption that these texts represent the constitution of a nation and should therefore mirror several aspects of the given society. However, the expansion of the corpus – if necessary - would have proved difficult, as there is an obvious limitation of the corpus size due to the few literary remains. A further problem encountered was the fact, that most literature taken into account is plain Christian or at least influenced by Christianity (c.f. “Beowulf”). This raises the question of reliability and objectivity, as most of the authors were clergymen, basically the only ones capable of reading and writing. Bede, for example, was a contemporary and it is therefore highly possible that some of the texts had been written in terms of proselytization instead of reporting the events impartially.
On the other hand, the literary remains are – apart from the archaeological ones – the only possibility to find out about a people which is no longer in existence. As there is no possibility to analyze speech via polls or transcripts, passive texts are the only way to gain insight from a linguistic perspective. Thus, the thesis has to be put up that even fictious literature mirrors – as nowadays – certain aspects of society. This thesis is supported by the fact that the discipline of cultural studies often uses literature as a source for its research.
3. Old English key words
3.1 Approaches to gain Old English key words
The problems mentioned above – especially the fact that there are only a few remains of passive texts – raises the question of how to find key words in a diachronic corpus. The issue is even complicated by a negative secondary effect of translation, the problem of synonymy; Old English terms might have several Modern English equivalents. Nevertheless, there are several possible approaches to find key words: The most promising attempt would be to check out general or basic concepts – thus gaining key words - of particular semantic fields, e.g. “nature”, which are provided online via the Google Trends tool. However, the tool delivers no basic terms but rather contemporary connotations, e.g. environmental problems as a key word for nature; thus, this tool as such will not prove sufficient for this research. Further approaches did not prove sufficient as well, for example the look for key words in OE glossaries, as these are written in Old English. The same goes for the key word function of Word Smith which does not work. Given these failed attempts, another approach has been taken: the compilation of a wordlist, which was searched and filtered for key words.
Structural words and unspecifying verbs as “(to) make” were obviously neglected when compiling the list of key words, as well as some types that are considered constitutive for a particular text form, e.g. the word “year” in a chronicle. The underlying thesis within this approach is of course that there is a relation between frequency and importance. The validity of this thesis is supported by several theories of text linguistics, especially the concept of cohesion. In this respect, words are considered important in a text when they are used multiple times. Hence, the first 315 types of the wordlist were checked, up to a frequency of 100 tokens. The key words thus gained were added up by further terms derived from them via semantical operations like defining synonyms, hyponyms and prototypes; in this respect, lemmas were used for most of the search terms, except when looking for the basic concept of a term; in this case the singular was used. The 15 most frequent OE Key terms are:
Abbildung in dieser Leseprobe nicht enthalten
Given these first results and the thesis mentioned above, it is possible to draw first conclusions concerning the structure of the Anglo-Saxon society: The favour of a class system combined with a strong hierarchy, the dominant role of the Christian church, the importance of warfare and the fact that the society was mainly patriarchal. On the basis of these key words, it is furthermore possible to allocate them to particular semantic fields and putting them into groups; thus establishing several domains that played an important role in Anglo-Saxon society. Certainly, the characteristics of these terms in context had to be analyzed first, as the characteristics of some words deviated from prior assumptions. The results of this qualitative analysis are to be found in chapter 4, whereas the following section will function as a quantitative analysis.
3.2 Semantic fields
This section is to be considered a pre-stage for the following quantitative analysis. As already indicated, the allocation of the key words to several semantic fields is possible. The respective fields with their subcategories are: politics (subcategories: court and government/ideals), religion (dignitaries/Christian concepts/Pre-Christian concepts), warfare, population (everyday life/family/special occasions) and nature (inanimate nature/animated nature). Having established these domains, it should now be possible to perform a quantitative analysis on the respective semantic fields.
3.3 Quantitative analysis
The arrangement of key words shows, that life in Anglo-Saxon society was determined by only a few specific domains; furthermore it is possible – taking the frequency- importance thesis into account – to hierarchize these fields concerning their importance for society. Considering the key words of each field, the total number of tokens indicates that Anglo-Saxon life was mainly influenced by Religion (6223 tokens), followed by Politics (4169), Nature (2400) and Warfare(1942). The field of “Population” got 5168 tokens in total, but cannot actually be seen as an influential social category as the key words are too general, not giving much information about societal issues. Yet the respective key words might prove a good source for quantitative analysis. The dominance of religion is not very surprising, given that fact, that spirituality played a much bigger role in former times. On the other hand, one also has to take into account, that most of the texts were probably written by clergymen. The influential status of politics is also self-explanatory, due to the inherently dominant role of monarchy, coinciding with authority, a class system and the need for a strong leadership. These results derived from analyzing the key words within the groups support the prior findings given in section 3.1. The relatively minor role of nature might be explained by the increasing influence of Christianity, thus eliminating a cult of nature which was characteristic for Pre-Christian concepts. The field of warfare got – surprisingly – the lowest number of tokens within the groups, yet given the fact that times of war normally represent a rather rare event in the course of life, the number is still relatively high and the effect on society cannot be neglected.
Within the field of Religion, one can clearly see that Pre-Christian concepts are on the vane. The whole category is dominated by Christianity whereas the concepts of polytheism and a middle-earth are vanishing. The subcategory dignitaries is a further proof for the importance of hierarchization, as those higher in rank are mentioned more frequently – except for the terms pope and archbishop. A reason for that might be, that on the one hand, archbishop is in terms of linguistics simply a modification of bishop; the reason that the pope is mentioned less frequently might be due to the fact that he was seated in Rome, whereas the bishops were the men in charge in Britain, hence having a higher influence and local power. The ideal of hierarchization is continued in the subcategory of Christian concepts; The term god and its equivalents as father and lord are used more frequent than those for Christ. Besides that, the frequency of antonyms, or at least semantically contrasting terms, might deliver information about society. For example, the term heaven is mentioned 5 times more than hell. This might indicate, that, in terms of proselytization, church rather focussed on the concept of hope than of fear and despair, which in a way contradicts our present view on church in the middle ages.
The ideal of hierarchization mentioned above is also present in the domain of politics; the more power a person has, the more frequent he is mentioned, from King (1588) to companions (78). Comparing the subcategories further supports the hierarchy thesis, as the ideals are less frequently mentioned than the titles for instance. Furthermore, many of the ideals do not represent the knightly codex of the late middle ages but rather coincide with the ideal of a strong leadership and hierarchy, such as glory, power and mighty.
With regard to the category of warfare, one can see that there is a tendency of individualization. That means, that single persons (hero, warrior) are brought out of the mass, e.g. the army, thus implicitly stressing their importance for warfare, furthermore idealizing them with the attribute of bravery. Whereas this attribution might confirm the present cliché of the strong and loyal warrior, the importance of a single man in battle might contradict our present view which assumed that battles were dominated only by the sheer mass of soldiers. Certainly, this picture of the outstanding hero might be influenced by the fictious aspect of some texts, nevertheless the assumption might not prove completely wrong when having a look at the frequency of these terms. A further fact that might contradict our view on Anglo-Saxon society is the relatively even distribution of the terms war and its antonym peace. This contradicts some clichés that see the Anglo-Saxons as a sheer brutal and violent people.
Information about the people in general can be derived from the group of population. Analyzing this group supports the assumption of a male dominated society, as the terms attributed to males are mentioned much more frequently. Interestingly, the term wife occurs even more often than the lemma wom*n, indicating that women are attached to an old gender stereotype, instead of being seen as an individual.
Within the last domain, nature, there is only one distinctive feature that is striking; the uneven distribution of summer and winter. The reason for this cannot be explained by a quantitative analysis, but by having a look at the key words in context, thus by examining their respective clusters and collocations. This qualitative analysis will be conducted in the following chapter.
4. Old English key words in context
When analyzing the context of the key word king, the cluster “this year died” (27) – especially due to the temporal deixis “this” - offers an interesting implication. As a certain year is characterized by the death of the king as an outstanding event, one could speculate that the life of the king could mark a particular period of time. The death and life of the king functioned therefore as a unit of time. Further clusters represent his status in society, his power and authority respectively. Within “the king sent” (7) and “the king ordered” (7), the verbs are, in terms of Pragmatics, performative verbs which mirror his might. On the other hand, the term king is not always marked in a pure positive, but sometimes also in an ambiguous way, as the cluster “the haughty king” (5) demonstrates.
In terms of clusters, earl stands in analogy to king. The cluster “this year died” (5) once again indicates, that the life of aristocrats functioned as a time span. Within the corpus, the term wise - in trivial concepts often attributed to kings – coincides with mankind in general, as the cluster “the wise men” (6) demonstrates, probably due to the grace of god; as a further cluster “wise of heart” (11) it also stands in connection to the bible.
Unsurprisingly, the term god is the most telling key word within the domain of religion, coinciding with many interesting clusters. First of all, “man of god” (36), a noun phrase used to describe almost exclusively clergymen and kings, but not the ordinary people, as to be seen in the concordance. Once again, this is an implication for a class system. This assumption is supported by two more clusters, referring to the same group of people: “servants of god” (15) and “beloved of god” (15). On the other hand yet, the cluster “dedicated to god” (8) does not refer to clergymen but mostly to virgins, thus probably embodying the Christian ideal of chastity. Besides, further clusters stress the importance of Christian concepts in contrast to the Pre-Christian ones. Within “the true god” (8), the definite article implies, as well as the modifying adjective, that the Pre-Christian concept of polytheism is not suitable anymore. In this respect, “word of god” is also mentioned 32 times, which stands for the attempt to convert the heathens, as to be seen in the concordance. The superior rank of god is shown by many collocates, such as the adjectives “almighty” (71), “holy” (39) and “eternal” (23). Inheriting the highest status demands also much devotion by the folk, as represented by the cluster “fear of god” (8), a fact that might have been abused by the king, as he is knowingly chosen by the grace of god.
Considering the term lord, a cluster analysis proves that the word is less used to denote an aristocrat – as might have been assumed - but rather to denote an equivalent for god: “lord of heaven” (26), “lord of men” (22), “god the lord” (16), “the lord god” (11). This thesis is confirmed by the collocates of Lord, that belong to the semantic field of divinity: “god” (83), “heaven” (51), “eternal” (33). Almost the same conclusions can be drawn for the term father – collocating with “almighty” (7) - whereas it is furthermore used as an equivalent for priest, as a typical cluster shows: „most reverend father” (10). With regard to the Christian dignitaries, the ideal of hierarchization is not only supported by the quantitative analysis but also by the collocates and clusters, as the terms of those higher in rank are modified by the respective adjectives, for example “the holy archbishop” (5) and “the holy pope” (7). The pattern “this year died”, already seen as marking the death of aristocrats as important, is also present with all Christian dignitaries: archbishop (16), pope (5), bishop (23) and abbot (5).
Further interesting facts are delivered by the terms monastery and church. Concerning the latter, there are two clusters to be found that are in need of explanation. The phrase “the primitive church” (6) might be considered pejorative at first glance, e.g. towards the natives. This assumption appears sound, as there has already been some evidence for the author´s negative attitudes towards Pre-Christian concepts. Yet, a check of the concordance demonstrates that this cluster simply refers to the beginnings of Christian Church in Britain when structures were less organised. The expression ”Custom of the [church]” (6) however implies a unifying thought established by this institution. Indeed, it refers to the universal or catholic church and its ideal of unification, thus once again fighting against Pre-Christian concepts such as a diversity of religions. Overall, the term church as denoting an institution is already established; within its context, there is an even distribution of using it either referring to the aforementioned or to church as a concrete building. Whereas the term church already embodies one of its present denotations, the description of the term monastery might deviate from our present views. It is to assume that monasteries are considered rather modest nowadays. However, in Old English texts, it collocates with the adjective “noble” (6). Yet, this adjective might of course also refer to the monastery as a pars pro toto for the whole institution or religion, which are certainly marked in a positive way. Hence, it might rather be the monastery as a substitute for an abstract term that is marked as noble, not the building as such.
A similar deviation from present assumptions can be identified in the semantic field of warfare. Given our general knowledge, one would legitimately expect that an adjective such as great should in first place collocate with terms as king or god. However, the cluster “a great slaughter” (11) ranks first, followed by “a great army” (7). With regard to this, the adjective great might rather function as a quantifier than as an expression of quality, thus further implying the importance of grandeur within this society. A look at the term army proves once again that special events were used to mark particular points in time - as implied by the cluster “year went the [army]” (20). Due to the post modification of the term army, the frequent phrase “the army that” (10) arose my interest as well. A look at the concordance line “[…] came against them with the army that they could collect” leads to several speculations about the quality of the army. It implies on the one hand that the army at times might not have been very professional as people could have been drawn involuntary, e.g. the thralls of the king. On the other hand, the verb to collect also implies that only volunteers might have been picked up. Anyway, the reputation of the army is therefore highly questionable.
Yet, a further example displays the potential might of the army, when looking at the concordance. The cluster “all the army”(9) appeared promising due to the premodifier “all”. Indeed, two concordancing lines led to speculations about the power of the army: ”Then the king and all his army proclaimed Sweyne an outlaw”, as well as “And all the army in East-Anglia swore union with […]”. The possession of political power as implied by these examples is of course questionable, but the soldiers appear here as coequals of the king. Thus, at least a strong company between king and companions is to be expected. As we know from history, battles were fought on fields, above that, the most frequent cluster of this key word “field of battle” (11) might add a linguistic proof to that. Interestingly, the frequent collocation of field and battle might lead to the speculation that warfare was much more important than agriculture. The further collocates of battle are in fact self-explanatory, such as war (11), warriors (9), death (7) and weapons (5). A linguistically interesting expression in terms of connotations is delivered by the composition “battle-grim” (4).The second component is actually negatively marked, yet it might stress the Anglo-Saxon´s pathos and bravery in this respect. The same goes for war, where “grim” (5) is also to be found as a collocate.
Those directly involved in war and fighting, the warriors respectively, were held in high esteem, as to be shown by the collocation with “hall” (8). The same is of course true for the heroes (5). The term itself already implies traits such as bravery, strength and valour, an assumption further enforced by the collocation with “hardy” (14). Besides, one could speculate if the rank of a hero is achieved by heritage instead of efforts. This might be assumed due to the collocate high (6), as it is related to high-born. Nevertheless the heroes are depicted as experienced warriors, proved by the adjective “hoary” (5). Experienced might have been gained from having slain many enemies. The verb forms slew and slain are very prominent in Anglo-Saxon vocabulary, certainly often in connection to enemy. Hostility between two opponents might have lasted a whole lifetime as two collocates of enemy – “old” (7) and “returned” (6) – imply. For the Anglo-Saxons, it seemed to be important by whom an enemy has been defeated, which is proven by the context of the keyword slain. Within its most frequent cluster “was slain by” (13), the agent is always mentioned, which is not that common for the passive form, but in this respect it might have been important in terms of honour. The collocates of the other verb form, slew, once again confirm clichés depicting the Anglo-Saxons as brutes, for example displayed by the cluster “flight and slew”. Hence, even when escaping, they killed enemies, or even worse, enemies were killed during their escape. The accusations are supported by other clusters; the phrase “and slew all” (7) occurs more often than “and slew many” (6). This might indicate that diplomacy was not much estimated.