Communicating Ideologies: An Examination of Web Definitional Examples

This study reports an analysis of definitional examples of country names posted on the World Wide Web (WWW). It aims at shedding some critical light on a sample of 33 definitional examples with an expectation of some sort of patterns which are indicative of ideological processes. The sample definitional examples were compared and conceptually labelled as positive, negative, and neutral within discourse analytic and Grounded Theoretic frameworks. Patterns were discerned and three categories identified: category I (European countries), category II (Asian, Russian, Middle & Far East nations), and category III (African nations). For the sake of objectifying the analysis, a list containing country names along with definitional examples was administered to 147 randomly selected university students to critically look at each definitional example and indicate the kind of image (positive, negative or neutral) the examples evoke in them. The analysis indicated that the definitional examples have the potential to communicate ideologies beyond exemplifying countries. Congruent to the analysis, images evoked in the respondents shifted from positivity to negativity as we moved from category I to category III countries. The study has brought further evidence for the claims of several studies that Africa is often negatively portrayed by Western Media.


Anecdote
It was almost 15 years ago. I was among a group of very young and ambitious college students gathered to discuss the possibilities of engaging ourselves to the service of our communities during an up-coming summer vacation. A high-ranking government official was in charge of the meeting. Issues of practical as well as theoretical nature were raised by the participants. The need to contribute to the development efforts of the country irrespective of our political views was succinctly presented by one of our friends. He was highly applauded when he argued that it is up to us to help the country prosper and change its images for the better mentioning that The Oxford Advanced Learner's Dictionary exemplifies the word famine using the phrase famine in Ethiopia.
Apart from adding to my motivation then to volunteer for the summer job in one of the economic sectors of the country, what our friend said remained in my mind and later came to suggest to me that it could be one of the elements from which people construct images of the country. The issue raised in this article was thus conceived some 15 years ago when I was a sophomore student and gave me the impetus very recently to undertake an examination of other definitional examples.

Objectives of the study
The above anecdote has brought an impetus for a closer look at some more definitional examples of country names. From previous research, we know that ideologies are embedded in texts (e.g., Hall, 1993;van Dijk, 2003;Bekalu, 2006).
As communicators, we therefore consciously or unconsciously produce and consume certain ideologies. Even though any text cannot be free from ideologies and hence the direct, explicit, denotative uses of words could bear ideologies of some sort (Hall, 1993;Eggins, 2004), ideologies tend to be subtler and stronger through the more implicit, connotative meanings of words. As such, through associations and connotations of words, certain positive or negative ideologies could be communicated across with varying degrees of impact on recipients. Ideologies communicated through texts of which we possess little or no prior knowledge appear to be more likely to have stronger impacts on us than the ones communicated through texts which we have already known something about. In this regard, ideologies communicated through definitional examples are more likely to have stronger impacts as readers or listeners would Online Journal of Communication and Media Technologies Volume: 2 -Issue: 1 -January -2012 normally consume such texts so as to learn words which they have little or no prior knowledge about.
We learn definitions and explanations of different words/subjects from a variety of print and electronic sources. For most of us across academic and professional spectra, definitions and explanations for a wide variety of words and/or subjects are available on the Internet. Owing to its advantages related to cost, time and mobility, most of us nowadays tend to find it easy and perhaps tempting to resort to the Internet to learn new things from. With the anecdotal datafamine in Ethiopiain mind, it is felt that more such definitional examples of country names, which most readers might be exposed to, 3. Is it possible to discern any sort of patterns in the ideologies communicated which are indicative of larger socio-economic and political systems?
In order to seek answers to these questions, data have been gathered and relevant analytic theoretical frameworks have been reviewed. I now start with the theoretical part.

Theoretical Underpinnings
Three issues have been taken as background for the present study: ideologies in texts, word meanings, and Grounded Theory. Below, a very brief review of the literature on these three issues has been made.

Ideologies in Texts
The notion of ideology is so elusive that it abounds with various definitions and conceptualizations. According to Durham and Kellner (2001:6), with a long history of several "abstract" conceptualizations and definitions, the term has for the first time been used to refer to everyday practices and rituals organized via social institutions designated as "Ideological State Apparatuses" by Althusser. An examination of recent literature also suggests that there seems to be consensus among scholars about the potential force of ideology for establishing different world views based on prevailing socio-cultural phenomena (Lassen et al., 2006). In recent works addressing the notion of ideology, we find Fairclough"s and van Dijk"s definitions. Fairclough (1992:87) defines ideology as "a construction of reality, which contributes to the production, reproduction or transformation of relations of domination." Within his socio-cognitive framework to discourse and text, van Dijk (2000:8) defines the term as "the basis of the social representations shared by members of a group." As exemplified by these two definitions, current scholarship conceives of ideology as a shared mental construct of a group of people derived from their respective socio-cultural contexts.
The process of producing and consuming texts entails what Fairclough (1995) calls discourse practices which in turn entail socio-cultural practicesthe social and cultural goings-on that shape the production and consumption of texts. In other words, what Fairclough calls discourse practices are the discursive ideological processes prevailing in a certain time among the members of a certain community. Such practices indeed result from socio-cognitive processes (van Djik, 2000) and shape our communication or the production and consumption of texts in Fairclough terms. As such, our communicative exchanges (what we speak, write, listen, read) are all reflections of our ideologies and in turn shaped by our ideologies. According to McCormick and Waller (2002), reading and interpreting a given text takes place in a situation where both the reader and the text possess literary and general repertoires which are derived from their society's literary and general ideologies. In fact, "Just as no text can be "free" of context (register or genre), so no text is free of ideology" (Eggins, 2004:10-11).

Ideology has become central in the interpretation of texts with a growing view of a text
as an open communication system where speaker and hearer or writer and reader come into contact with their own distinctive, individual interpretations (Tilborg, 1991).
Though every text is an embodiment of certain ideologies that can be critically studied and identified (Stillar, 1998), the focus of several contemporary studies have been on media texts owing to the presumed power of such texts to reach and "influence" a huge number of people (e.g. Trew, 1979a;1979b;Hodge, 1979;Van Dijk, 1988;O"Halloran, 2003;Bekalu, 2006). Alongside its growing use and advantages related to cost, time and mobility (Griffiths et al., 2006), the World Wide Web (WWW) has long become the information super highway for most of us. The World Wide Web has also increasingly become intermingled with traditional mass media, and as such its products are worthy of investigation.

Word Meaning
The philosophical question of what is word meaning has long drawn the attention of scholarships across the humanities and sustained heated debates (Bouillon and Busa, 2001). All the inquiries, however, agree on the fact that words of any language can have different shades of meanings. Linguists and language philosophers have various ways of theorizing about these differences: semantic vs. pragmatic, literal vs. figurative, explicature vs. implicature, denotative vs. connotative, etc. The former types in each pair refer to the direct, dictionary definitions of words whereas the latter are often referred to as context-sensitive, functional, personal or emotional meanings (Carter et al., 2001). While members of a language community tend to readily agree on the former types of meanings in each pair, considerable differences occur among people with regard to the latter types in each pair. According to Carter (2001), many people are likely to agree on implicated or connotative meanings of words when the implicatures or connotations are strong. Such words can evoke strong negative or positive connotations and can have a powerful emotional impact (Carter, 2001:102).
Semiotics posits that ideologies of speakers or writers often get their way through connotations or implied meanings (Noth, 1990). In this case, ideology is seen as a sort of second-order semiological system, which can be characterized at the level of connotation (secondary language) as opposed to the level of denotation (primary, Online Journal of Communication and Media Technologies Volume: 2 -Issue: 1 -January -2012 ordinary language) (Larrain, 1979). In his influential Encoding/Decoding model, Hall also states that we can see more clearly the active interventions of ideologies in and on discourse at the level of connotation. However, Hall (1993:512) remarks that "this does not mean that the denotative or "literal" meaning is outside ideology. Indeed, we could say that its ideological value is strongly fixedbecause it has become so fully universal and "natural"." According to Hall (1993:512), "the terms "denotation" and "connotation", then, are merely useful analytical tools for distinguishing, in particular contexts, between not the presence or absence of ideology in language but the different levels at which ideologies and discourses intersect." In short, based on the available literature, it appears to be possible to conclude that ideologies are always embedded in our words and tend to be working more at the level of connotations than denotations. Critical readers searching for ideologies need to attend to both levels of meanings.

Grounded Theory
The present study began with a critical look at and explication of implied/connotative The notion that inquiries seeking ultimate answers to the longstanding human quest for truth might not be adequately addressed within the narrow boundaries of modernity or positivistic scholarship has apparently led to a paradigm shift to post-modernity where Grounded Theory (Glaser and Strauss, 1967) with its focus on qualitative data and inductive procedures has prevailed. Grounded Theory posits the inductive discovery of theory grounded in systematically analysed data. Deriving its tenets from the bigger philosophical notions such as pragmatism and symbolic interactionism, Grounded Theory advances the systematic analysis of data leading to a theory that fits at least one set of data instead of grappling with theories that may not fit any data at all. The theory is built on compared concepts; that is, similar data are grouped and conceptually labelled, concepts categorized, categories linked and organized by relationship, conditions and dimensions developed, and finally a theory emerges (Scott, 2004). Put in other words, the researcher analyses the data comparatively and concurrently and identifies core categories or variables which, according to Byrne (2001), will serve as the foundation for theory generation. The core variable is characterised, among other things, by recurrence and links various data and also has an explanatory function towards theory formulation (Byrne, 2001).
The destination of all Grounded Theoretical inquiries may not indeed necessarily be the formulation of theory proper. As a matter of fact, most studies within this theoretical framework build in a systematic manner from specific data to some sort of generalizations. Several studies ranging from pure sociological inquiries (where the theory traces its origin) to a wide array of social sciences including accounting, business management, education, nursing, public health, and social work have used Grounded Theory (Strauss and Corbin, 1997).
In short, the present study has been conducted against the backdrop of the three theoretical issues outlined aboveideologies in texts, word meaning, and Grounded Theory. It is felt that texts (whether spoken, written or visual) are embodiments of ideologies which are reflections of our socio-cultural, economic and political environments. Ideologies are embedded in our words and are accessible to a critical reader/listener who attends to not only the direct, literal meanings of words but also to the implied, connotative meanings. In line with Grounded Theory, it is also believed that some generalizations that would probably lead to theorizingtogether with other similar existing and future inquiriescan be systematically built from anecdotal, small data.

The Data
Definitional

Analytical Framework
Among the various analytical approaches, discourse analysis has been chosen for its potential to deal with the subtle links between texts and ideologies. Indeed, the approach has long been chosen as a fitting analytic tool for scholarships pursuing the notion that there is no "value-free" science (van Dijk, 2003) and set out to assess the multifaceted subtle links between texts, ideologies and larger socio-cultural, political and economic structures. As opposed to other analytical tools which are more focused to the interplays between or among "micro textual elements", discourse analysis affords analysts the possibility to map out and pinpoint interactions of micro textual elements with the larger (macro) discourse and socio-cultural systems. In this study, therefore, Fairclough"s (1995Fairclough"s ( , 1998 analytical framework has been used. Fairclough (1998) advances a framework of critical discourse analysis which is particularly suited to political discourses. In this framework, three levels of analysis have been put forward: text, discourse practice, and socio-cultural practice. Fairclough (1998:144) briefly describes each sort of analysis as:  analysis of texts (spoken, written, or involving a combination of semiotic modalities, e.g., televisual texts);  analysis of discourse practices of text production, distribution and consumption; Online Journal of Communication and Media Technologies Volume: 2 -Issue: 1 -January -2012  analysis of social and cultural practices which frame discourse practices and texts.
According to Fairclough, the link between texts and socio-cultural practices is mediated by discourse practices. These three sorts of analysis can also be described in terms of sub-or superset systems. Texts are a subset of discourse practices, which are in turn a subset of socio-cultural practices.
The present study attempted to analyze the sample definitional examples at these three levels. Through the analysis of the definitional examples (texts), it has been attempted to shed light on the discursive ideological practices that have helped produce the texts.
The impacts of prevailing global socio-cultural practices on the production of texts through discourse practices have also been discussed. What is more, patterns in ideologies have been discerned through comparisons of the data and the identification of core categories (variables) informed by Grounded Theory.

Texts here are definitional examples of country names. Consistent with Grounded
Theoretical procedures, the data (definitional examples of country names in our case) have been compared and categorised (Scott, 2004). They then have been conceptually labelled so that core categories could be identified. Informed by insights obtained from with arguably positive images have been noted in this category. Figure 1 summarises the inductive analysis of the data informed by Grounded Theory. Now, let us see the discursive and socio-cultural practices within which these texts have been produced and consumed.

Objectifying the Analysis
As already mentioned, a list containing country names along with definitional examples has been prepared and administered to 147 randomly selected university students to critically look at each definitional example in each category and indicate the kinds of images they evoke in them. Three kinds of imagepositive, negative or neutralwhich the analysis identified as core categories have been presented against each definitional example. In doing so, an attempt has been made to make the analysis and the entire study for that matter as objective as possible. The data obtained from student ratings have been presented in three tables (see below) and the results have been summarized as follows:

Results
As can be seen in table 1, the large majority of Category I countries evoked positive and neutral images in the respondents. When we look at the mean percentages, whereas 48.8% and 41.9% of the respondents reported positive and neutral images respectively, only a token (9.4%) reported negative image. Comparatively, the countries in this category evoked more positive than neutral images. When we come to Category II, a large majority of the respondents (58.3%) reported neutrality, whereas a considerable percentage (28.2%) of them reported positivity, and only a small percentage (13.5%) of them reported negativity (see table 2). But still, comparatively, countries in this category evoked more negativity than countries in the first category. With regard to Category III countries, a great majority (64%) of the respondents reported negative images followed by a small percentage (26.5%) of neutrality and a tokenistic (9.5%) percentage of positivity (see table 3).
When we look across the three categories, positivity decreases as we move from Category I to III. On the contrary, negativity increases as we move from Category I to III, and no pattern can be discerned with regard to neutrality. It should be noted that these findings are congruent to the findings of the analysis discussed above.

Concluding Remarks
Inspired by anecdotal data, the inquiry reported here has led to some generalizations. In line with the principles of Grounded Theory, an overall inductive procedure has been adopted. The data (definitional examples) have been collected, compared, categorized and core categories (variables) have been identified. The data obtained from the sample students about the kinds of image the definitional examples evoked in them helped to objectify the analysis and the entire study. The core categories which have been labelled as positivity, neutrality and negativity, have been found to display some of their professed characteristics (Byrne, 2001). That is, they displayed recurrence across groups of countries and linked countries of similar socio-cultural and economic structures to one another. For instance, while positivity and neutrality linked European countries, negativity has been found to be the bond among the majority of African countries. The links between texts (definitional examples) and ideologies have been discussed within Fairclough"s analytical framework.
The analysis made within Fairclough"s framework has shown that micro textual elements (words) which seemed to have been used only for exemplification purposes could be positively or negatively loaded, as in the case between the "Austrian music" vs.
the "Angolan Civil War", suggesting the presence of subtle ideological processes. As writers or speakers, we normally choose words that best communicate our meanings unless we intentionally want to make our ideas obscure or ambiguous. And we always have several ways of saying the same thing. The very choice of the word "music" to exemplify the word "Austrian" and "civil war" to exemplify "Angolan" shows the kind of ideologies the writer(s) possess concerning at least these two countries and at most the continents the countries come from. Ideologies are shared mental constructs of a group of people derived from their respective socio-cultural contexts. As such, the production of such texts could be taken as one instantiation of the discursive practices of the writer/s" community in particular and the larger global socio-cultural environment in general.
The analysis supported by the objectifying data has shown patterns with regard to positivity and negativity across the three categories of countries. The depicted patterns appear to mirror the existing socio-economic and political situations of the countries.
This further strengthens the argument that these definitional examples are not just examples but are indeed manifestations of subtle ideological processes. Positivity and neutrality do not seem to be problematic, when we look at them from ideological point of view. Negative portrayal, however, deserves some sort of scrutiny for the simple reason that it is not at least something that anyone would prefer to be associated with.
The definitional examples for these negatively portrayed countries might have been chosen by their writers for their potential to show the situations in the countries they refer to as the writers see them, informed by their ideologies. However, the portrayals are indeed at odds with what the countries would like to be known for as can be gathered, for instance, from the web posts (defining features) of the countries" foreign ministries. It goes without saying, for example, that Ethiopians or at least the country"s government, as can be gathered from several pages of the country"s foreign ministry website, would be happier to see their country associated with its historic obelisks (Axum) and the rock-hewn churches of Lalibela, among other things, than their "immigrants". The present study brings further evidence to the much debated issue of the negative portrayal of Africa by the Western media (e.g. Palmer, 1987;Zaffiro, 1992;Hagos, 2000;Gilkes, 2000). All these writers contended that western media portray Africa negatively. Zaffiro (1992:81) vehemently contends that the "erratic, unbalanced or missing coverage" of Africa by the western media "creates and reinforces dangerously ahistorical, inaccurate, negative cultural stereotypes." Congruent to these arguments, the association of African countries with such negatively loaded words as "civil war", "desert", and "terrorism" in definitional examples supposed to be read by people who are learning about the countries for the first time has the potential to mislead people to Online Journal of Communication and Media Technologies Volume: 2 -Issue: 1 -January -2012 bear biased images. In general, the "lopsided" system of contemporary international communication (Fortner, 1992) seems to have put the West in a situation where it can speak for the rest of the world, especially the Third World, with little or no means of counter-action available for the Third World.
While it has well managed to identify and draw attention to the subtle links between text and ideology in general and the processes of communicating ideologies through web definitional examples in particular, the present study has its own limitation, the awareness of which could lead to further research. The main limitation of the study is related to the fact that respondents from whom objectifying data have been solicited are all from the same country with more or less similar socio-cultural and economic background. Even though their responses have been congruent to the analysis performed by the author, the question whether or not the same kind of responses could have been obtained had the respondents come from a different socio-cultural and economic background remains unanswered and might well be an issue for further research.