A COGNITIVE ANALYSIS OF EPISTEMICALLY-MARKED DISCOURSE (WITH SPECIAL REFERENCE TO ADVERBS OF OBJECTIVE EVIDENCE)
A COGNITIVE ANALYSIS OF EPISTEMICALLY-MARKED DISCOURSE (WITH SPECIAL REFERENCE TO ADVERBS OF OBJECTIVE EVIDENCE)
Abstract
The paper presents a cognitive analysis of epistemically-marked discourse, with adverbs of objective evidence as an essential constituent. The number of adverbs selected for the examination (evidently, obviously, apparently, clearly) is extended by adding naturally, which, though not acknowledged by many linguists as an evidential adverb, is enlisted here due to a number of indicators showing its semantic closeness to the target group of adverbs. In the discussion concerning the distinction between epistemicity and evidentiality, it has been suggested that these two categories are not in equal relation to each other, epistemicity standing in the hypernymic relation to evidentiality. The analysis is based on two cognitive concepts: evidential justification and epistemic support. In case of the absence of evidential justification, the information is provided through vertical context. The paper offers a further division of this concept into two types:
1) intravertical context the needed information is drawn out from larger fragments of text;
2) extravertical context – the needed information is extrapolated from outside the text.
1. Introduction
1.1. Discourse vs text
Linguistic studies have long shifted from analyzing the sentence as the basic syntactic unit to studying aspects of language beyond the sentence. Linguists have come to the conclusion that language cannot be studied in isolation from the communicative intentions of language users and the context within which language is used . Hence the concept of discourse came into being, which is inherently linked with the name of Z. Harris , who defines it as “a method of seeking in any connected discrete linear material some global structure … characterizing the whole discourse or large sections of it”. L. Hoye views discourse as “all aspects of language organization (whether structural or not) that operate above the level of grammar”. H.Widdowson points out the basic function of discourse analysis as the operation of language code in stretches of text larger than the sentence. It becomes clear that the definition of discourse is invariably established through text or reference to it. Not surprisingly, the terms text and discourse are often used interchangeably, causing considerable confusion in defining their concrete meanings and functions. Therefore, we sometimes come across overlapping definitions of these concepts, positing them often at the same level. However, many linguists hold that these two categories have distinctive features in terms of organization and communication. Brown and Yule see the difference in terms of cohesion and coherence, pointing out that coherence means cohesiveness and understanding between units in a text or utterance, whereas cohesion implies organizing the inner connectedness between propositions to get wholeness. Some linguists view the difference between text and discourse in the framework of functional sentence perspective. According to Van Dijk , text and discourse differ from each other in that the former is related to language competence, whereas the latter to language use. Brown and Yule distinguish text as a final product and discourse as a process of text. Other linguists Berrendorren ; Cornish ; Avanzi establish a distinction between text and discourse in terms of ‘micro-syntax’ and ‘macro-syntax’, relating them to text and discourse respectively.
The most rational approach seems to be demonstrated by L. Hoye who claims that text-discourse distinctions are not only inconsistent among themselves but also untenable, pointing out that the distinction reflects a basic difference of emphasis. He suggests studying these two categories together: firstly, to relate them in terms of a general set of semantic categories (e.g.cause-effect) and secondly, to examine sentences as products of actions performed by speakers or writers. Based on the approaches adopted by linguists, we can conclude that discourse means the unity of text and context, whereby context plays an essential role in the global understanding of the text, and this will be the main guideline of our research.
A text has a unity of structure, which is provided by various means, such as morphological, syntactic, lexical, pragmatic, and stylistic. Among morphological means, a considerable place is given to epistemic adverbs, which is the object of the present research.
1.2. The nature of epistemic modal relations
One of the main purposes of human interaction is to convey information as well as to express the attitude of interlocutors towards it. Linguistically, this function is predominantly performed through the category of modality and finds its reflection in many devices, including modal words (primarily verbs and adverbs) that almost all languages of the world have at their disposal. The category of modality distinguishes two main types of modality: deontic and epistemic. Let us look at the following examples.
A. He spoke naturally and assuredly at the meeting. (deontic modality).
B. He naturally refused to speak against his friend. (epistemic modality).
Considering the different uses of naturally in these two sentences, we can see that in (a) naturally functions as an adjunct, i.e. integral part of the structure serving as an adverbial modifier of manner; in (b) naturally is a disjunct, not being integrated into the structure, acting as parenthesis and having only a communicative function.
Linguists acknowledge two levels of linguistic dependency: syntactic and semantic. According to P. Pietrandrea , epistemic modal adverbs always govern their scopes semantically (by predicting their truth values) but are syntactically independent of them. Indeed, naturally in the above example (b) is not integrated into the structure of the sentence, its role is only to present the utterance as a fact, i.e. has an illocutive force. The same utterance containing another adverb, for example, in He, fortunately, refused to speak at the meeting, “fortunately” conveys quite a different illocution (approval, satisfaction).
1.3. The place of evidential adverbs in various classifications
The present analysis is aimed at studying the epistemic markers such as evidently, obviously, apparently, and clearly. In linguistics, they are acknowledged as evidential adverbs, the basic characteristic of which is the expression of the speaker’s assessment of proposition information based on evidence or drawn from it. We assume that the number of evidential adverbs can be extended by including naturally. We proceed from the fact that the definitions of natural recorded in different dictionaries show that they coincide on many points with those of apparently, evidently, obviously, and clearly For illustration, we can bring their definitions given by OED: obviously – in a clear perceptible manner, evidently, plainly, naturally, as might be expected from the circumstances; apparently – evidently or manifestly to the sight, visibly; evidently – to be seen or understood easily; clearly: in a way that you would expect; as a normal, logical result of something. naturally – in a way that is very obvious, as a normal, logical result of something. As is seen, the target adverbs are often defined in terms of other adverbs of the same group and the definition of naturally is associated with such adverbs as obviously and clearly. It is worth mentioning that M. Carretero et al. , who are closely involved in examining evidential adverbs, note that “the sense of naturally, as might be expected from the circumstances, is very often found with obvious”. Besides, the analysis of our factual material confirmed the fact of the semantic and pragmatic closeness of naturally to the adverbs of the target group. Thus, we think that there seems to be enough ground to list naturally among the adverbs of objective evidence.
1.4. Epistemicity vs. evidentiality
It is important to make it clear what relation epistemicity and evidentiality are to each other. We assume that the current distinction between these two concepts is grounded in the wrong approach to the nature of the relationship between them. The controversial view consists in considering them as belonging to different layers in the hierarchical structure. We proceed from the fact that epistemicity stands in the hypernymic relation to evidentiality. In other words, evidentiality enters the scope of epistemicity as part of a lower level; consequently, all types of sentence adverbs, evidentials included, should go under the general term epistemic. However, many researchers use the terms evidentiality and epistemicity indistinguishably. K. Boye unites these two categories under a common umbrela, though he regards them as parts of the distinct and largely independent conceptual domain. P. Roseano et al. claim that the relationship between evidentiality and epistemicity has been approached differently by various authors, and most researchers nowadays support “an in-between approach, whereby the epistemic and evidential categories are interrelated and overlap to a certain extent, yet are neither completely conflated nor may be treated totally separately”.
As most linguists note, the basic characteristic of evidential adverbs consists in expressing the speaker’s evaluation of information based on objective factors such as direct observation or inference. As it is fixed in the Concise Oxford Dictionary of Linguistics by P. Matthews , the basic difference between evidential adverbs and other types of epistemic adverbs consists in the fact that “the meaning of evidential adverb is closely connected with the source and reliability of the evidence on which a statement is based”. According to F. Haan’s definition, “evidentiality refers to the marking of the source of the information of the statement, while epistemicity refers to the degree of confidence the speaker has in his or her statement”. As it can be inferred from the quotation, epistemicity, in the author’s interpretation, should be understood as referring to attitudinal adverbs (certainly, surely, perhaps, etc.).
E. Traugott examines the difference between the evidential group and other types of epistemic adverbs in terms of subjectification and intersubjectification. Epistemic adverbs such as certainly, probably, possibly, surely, perhaps, and the like, express the speaker’s personal knowledge as to the truth value of the proposition. Therefore, they are seen as affected by subjectification, i.e. the speaker’s evaluation, while evidential epistemic adverbs are distinguished by intersubjectification as they rely on objective factors.
All this comes to prove that evidential adverbs form a separate class of lexical units which by their semantic and pragmatic meanings are distinguished from the other types of epistemic adverbs. It is this group of epistemic adverbs that is the object of our study. There are a number of interesting researches made in this field : Merlini, Barbaresi ; Vandenbergen and Aijmer ; Roseano et al. ; Musi and Rocci et al. , but they are made basically within the framework of argumentation theory. The present research is based on the cognitive theory, K. Boye’s theory in particular, which allows us to conduct a comprehensive and multifunctional analysis of an epistemically-marked discourse.
As illustrative material, we have chosen literary works by American and British writers. This choice can be explained by the fact that the tasks of discourse analysis require larger contexts, and in order to render all possible judgments and justifications in the uses of a particular epistemic adverb, we sometimes have to deal with not only one or two sentences, but several paragraphs, chapters or even the whole book – something that corpus cannot presumably provide.
2. A functional analysis of epistemically-marked discourse
2.1. Horizontal context
Cohesion and coherence are two important categories that are very important for discourse analysis. M. Halliday and R. Hasan hold that “the concept of cohesion is a semantic one; it refers to relations of meaning that exist within the text, and define it as a text”. But cohesion is not sufficient to form a coherent discourse. D. Crystal defines coherence as “the main principle of organization which is assumed to account for the underlying connectedness of a piece of language”. The same idea is expressed by Van Dijk. In his book Text and Context, the linguist says: “Coherence is a semantic property of discourse, based on the interpretation of each individual sentence relative to the interpretation of other sentences.” The author claims that the coherence of discourse is represented on two levels: linear or sequential and global. Linear coherence refers to “coherence relations holding between propositions expressed by composite sentences and sequences of these sentences. Global coherence is of a more general nature and characterizes a discourse or its larger fragments as a whole.
This part of the research is devoted to the analysis of evidential adverbs at the level of linear or horizontal coherence. In this respect, we would be guided by the cognitive theory of K. Boye , who holds that the category of epistemicity is composed of two subcategories: evidential justification and epistemic support. Evidential justification is the mention of the evidential source that justifies the validation of the truth of a linguistic representation, i.e. epistemic support. They correspond to two parts of epistemically-marked discourse (EMD) on the cohesive level: scope constituent (SC) and epistemically-marked constituent (EMC). It is obvious that these two subcategories are closely linked with each other by various semantic, syntactic, and pragmatic ties and it is impossible to have one without the other. Let us look now at the following example.
1. Naturally, his death meant very little to me at that time – there were eighty deaths in my own battalion on the day George was killed, and the Armistice and setting my own problem again and getting to work, all occupied my attention .
Taken in isolation, the first clause Naturally, his death meant very little to me at that time sounds extremely inhuman, even sacrilegious. But the succeeding sentences explain the psychological and physical state of the person who produced that utterance, a man who saw deaths every day and was obsessed with so many strenuous and time-consuming tasks.
Evidential justification is based on direct observation realized through mental or physical perception. Mental perception is the speaker’s ability to draw inferences from the proposition providing evidential justification for the use of an appropriate evidential adverb. It is characteristic of all the target adverbs, irrespective of the position they occupy in the EMD. Consider the following utterances, in which the evidential justification is in bold.
2. The evidence showed that the deceased lady, while attempting to cross the line, was knocked down by the engine of the ten o'clock slow train from Kingstown, thereby sustaining injuries of the head and right side which led to her death… Evidently she has been unfit to live, without any strength of purpose, an easy prey of the wrecks on which civilization has been reared .
3. I took Strickland’s temperature. It was a hundred and four. He was obviously very ill .
4. Naturally Harvard had wanted George to stay on. The department chairman even called in Kissinger to discuss how they could persuade the young scholar to remain in the academic ranks. His adviser countered that George was a strong-willed man .
5. Apparently she also liked him. And Mrs. Lamberton noticed it with that swift rather devilish intuition of women .
6. For the present he seemed quite lost. When I suggested that he should go to bed he said he could not sleep; he wanted to go out and walk about the streets till day. He was evidently in no state to be left alone .
The evidential justification can be the product of inferences drawn not only from conceptual perception as illustrated above but also from sense perception (sight, hearing). It is logical to think (the contextual analysis confirms it) that the SCs based on sensory observation provide a more trustworthy and reliable justification compared with the SCs based on reasoning. Consider the following EMDs in which the underlined SCs highlight the evidential justification based on sight and hearing.
Sight
7. As she approached the house he was surprised to see a state police van parked in front of her house. Evidently the crime scene investigators were still there .
8. And swinging wildly, he proceeded to chase a turkey – which apparently only he was able to see – around the perimeter of the dining room .
9. … but they obviously formed a self-sufficient little group, and once their umbrellas, bamboo rugs, dogs, and children were set out in place part of the plage was literally fenced in .
10. Clearly, everyone had left in a hurry. The sofa was opened out. The kitchen table was littered with dishes and no one had put the cream away .
Hearing
11. She was evidently a copious talker, and now poured forth a breathless stream of anecdote and comment. She made the conversation we had just had seem far away and unreal .
12. “With your stature in the town a word from you would go a long way in getting them to do something,” said Angela. Sherwood trumped forward in his chair. He was clearly flattered .
13. Somebody was obviously living here. Sophie could hear wood crackling in the old stove. Someone had been here very recently .
On the basis of the illustrated examples, we can assume that evidential justification is distributed among the SCs in the following way: conceptual perception – all the target adverbs, with naturally and obviously prevailing; sight – evidently, apparently, obviously; hearing – evidently, clearly. The distribution of evidential adverbs according to their reference to conceptual perception, sense and hearing is represented in Table 1.
Table 1 - The distribution of evidential adverbs
Evidential adverb | Conceptual perception | Sight | Hearing |
Evidently | + | + | + |
Apparently | + | + | – |
Obviously | + | + | + |
Naturally | + | – | – |
Clearly | + | + | + |
As is shown in Figure 1, naturally is not observed in the SCs based on sight or hearing; nor is evidently found in the SCs based on hearing.
2.2. Vertical context
This part of the research is devoted to the analysis of an EMD at the level of global coherence, which is of a more general nature and characterizes a discourse or its larger fragments as a whole . This characterization is congruent with Allerton’s definition of this notion: “Discourse analysis is a method of seeking in any connected discrete material … some global structure characterizing the whole discourse.” Viewed from the micro-syntax and macro-syntax theory, discourse is associated with macrostructure. In this respect, we would cite F.Cornish , who says: “A purely grammatically-determined (micro-syntactic relation) has none of the implications associated with macro-syntax. Beyond a given sentence, we enter the realm of intersentential relations, which is the province of discourse (and hence of “macro-syntax).” N. Fairclough supports this notion by saying: “There is a set of other texts and a set of voices that are potentially relevant, and potentially incorporated into the text”. This thesis is congruent with Brown and Yule’s viewpoint which emphasizes the importance of participants’ ‘backward knowledge’ in the interpretation of discourse coherence stored in memory, taking such forms as frame, schemata, script, scenario, and plan.” O. Akhmanova and I. Giubbenet view this concept as vertical context which they see as a set of information of historical, geographic, cultural, and pragmatic nature, without which genuine understanding and interpretation of the text is impossible.
We will enlarge the borders of this concept and include the data received not only outside the discourse but inside it. Consequently, we can speak of two types of vertical context which we can term intravertical context (the data is accumulated from the content of the global text) and extravertical context (the data is installed beyond the text).
The notion of vertical context has proved applicable in examining the cases with the EMD when evidential justification is not found within the immediate context but extrapolated across the larger fragments or even through the whole text (book).
Intravertical context
14. George Keller had resigned himself to eating lunch alone on a courtyard step.
Clearly, no one near or dear to him was that day present .
The absent information (why eating lunch alone?) is drawn out from the content of the novel to justify the use of the evidential adverb clearly. We find out that it had been the practice for three centuries to hold the Harvard university’s Commencement Day when the graduates receive diplomas in the atmosphere of great solemnity and festivity, with their parents and close friends proudly present at the ceremony. But George was a Hungarian who fled from his country to become an American citizen and though having achieved great success in his somewhat dubious career, he could not realize himself as a genuine American citizen and kept himself aloof from his classmates, let alone his family with whom he had long lost contact. These circumstances account for his solitude on that memorable day.
Consider another example.
15. Actually the publicity had bothered Hess, who had been convinced that political pressure would cause the Government to back down on enforcement of the Immigration Act, as had happened many times before. Apparently, though, he had been wrong .
We will briefly describe the case. The event takes place in Canada. It is connected with a man named Duval, who wants to get a Canadian visa and is always rigorously refused. But this time he wins. That Hess was wrong in his judgment concerning the Canadian government’s immigration policy, which becomes known to the readers only at the end of the book when they read the following lines uttered by the Prime Minister: “I issue a press statement this afternoon that Duval will be given a temporary immigrant visa at once… Also, on my personal recommendation, the Cabinet will consider an order in council allowing Duval full immigrant status as quickly as possible” .
Extravertical context
16. All afternoon Danny listened spellbound as his classmate cascaded with ideas. Naturally, they couldn’t cram Joyce’s whole epic novel into two hours of stage time, but they could concentrate on the “Nighttown” episode, when the protagonist, Leopold Bloom, wanders through exotic parts of the city .
If the readers are not acquainted with James Joyce’s book” Dubliners “(it is a series of splendid stories about middle-class inhabitants of Dublin), it would be difficult for them to understand why Danny and his group could not transmit the book’s magic spirit in two hours’ time performance.
Consider another example.
17. Naturally, being Kissinger’s shadow, George not only held opposing views but was actively involved in the escalation of hostilities .
To understand why George blindly followed Henry Kissinger’s views and actions, it is necessary to know what kind of political figure the latter was. According to David Greenberg, professor of history at Rutgers University and author of “The Republic of Spin and In Nixon's Shadow”, Henry Kissinger is one of America’s “most controversial and important diplomats in all his dimensions – as a serious geopolitical thinker, a skilled bureaucratic infighter, a relentless courtier of power, and a virtuoso of self-promotion” . Without this information obtained through extravertical context it would be difficult to interpret George’s blind imitation of Kissinger’s behavior, himself being trustworthy enough to be included in almost all strategy actions undertaken by his chief.
Thus, we see how significant is the role of vertical context in the deep and comprehensive understanding and interpretation of the text.
3. Conclusion
The cognitive study of epistemically-marked discourse offers to some extent a new approach to the study of evidential adverbs. The analysis is based on two cognitive concepts that characterize the relations between the constituents of the epistemically-marked discourse: evidential justification and epistemic support. In case of the absence of evidential justification, it is installed through the vertical context, i.e. extralingual factors such as the speaker’s / the reader’s knowledge of the world, their education, learning, and life experience. Two types of vertical context have been established in the paper:
1) intravertical context – the needed information is drawn out from the larger context;
2) extravertical context – the needed information is extrapolated from outside the text.
The number of evidential adverbs selected for our examination (evidently, obviously, apparently, clearly) is extended by adding naturally, a seemingly controversial choice, but its inclusion has been substantiated by a number of points showing its semantic closeness to the target group of adverbs. In the discussion concerning the distinction between epistemicity and evidentiality, it was proved (unlike some other pieces of research) that these two categories are not in equal relation to each other, epistemicity standing in the hypernymic relation to evidentiality.