The meaning of a term is determined by a definition. Definitions are, therefore, conventions of a language. Terms are words that describe, for example, objects, processes, characteristics of objects or persons, or notional content of our imagination.
The language that we learn forms the foundation for every definition. Terms are attributed to phenomena, which science investigates, if our day-to-day language is capable of describing them, however inadequately. Definitions enrich and concretize the basic vocabulary of our day-to-day language. This language consists of words the meaning of which we learn essentially during childhood and adolescence. Words that possess content are linked by recurrent attribution of words to phenomena (objects, processes, and ideas). In this manner, we learn the general usage of day-to-day vocabulary. We learn the meaning of scientific terms not by recurrent examples; we determine their meaning through known words from the day-to-day language by means of a definition.
Because various disciplines of science are specialized and go deep into the core of every subject, a conceptual differentiation is necessary. For this purpose, new words are coined or borrowed from other (living or extinct) languages. Naturally, even words from the day-to-day language are attributed to scientific terms, but, in most cases, their meaning must be redefined in order to achieve the required degree of precision. Since the language of science also aims to facilitate proper comprehension, there must be unanimity about the usage of the terms. If new or inaccurate terms are introduced without a prior definition, their meaning will remain ambiguous or completely obscure. Nevertheless, comprehension is also lost when new definitions are formulated instead of using established terms or simple descriptions.
Nominal Definitions
Nominal definitions interlink terms (nouns) – nouns are attributed to other nouns. That is to say, a word is defined as a term by equating it with other known words. The meaning of a term (definiens) to be defined is determined as synonymous with the description of a fact through known words (definiendum). For example, “mass media” (definiens) are defined as “institutions that periodically present communication content to an open public”) (definiendum). This example makes the problem of nominal definitions clear: the usefulness of a nominal definition depends on the usefulness and the clarity of terms with which the definiens is determined. Also, the terms of the definiendum would have to be defined (in the example, particularly: “institutions,” “communication content,” and “open public”). Any attempt to give a “complete nominal definition” is doomed to fail, because terms are always explained with terms that, in turn, need to be defined through terms. This problem is called an endless definitional regress. Nevertheless, as a rule, such chains of definitions can be broken if we can assume that the terms of definiendum are generally known.
A definition becomes useless if it is circular. Circularity is always present if the definiens is contained within the definiendum. When terms of the definiendum are later defined in stages, the risk of hidden circularity increases. To illustrate this: a “journalist” is defined as a “person who possesses a journalist’s identification”; in the next step, a “journalist’s identification card” would be defined as a “document to identify a journalist”. In principle, nominal definitions cannot be incorrect or false; they can only be either useful or useless. Moreover, there must be certain unanimity about the usage of terms. Normally, generally known terms from the day-to-day language or established terms from the disciplines of science are analyzed for their usage (explicated) and, if necessary, more precisely defined. Whenever a term occurs even in the natural language or, as a rule, if it is already established in the discipline, it must be verified whether the given definition covers all that it is expected to cover. Only when a new object or phenomenon is described and a new term is coined for it, an absolute sovereignty of definition is present with the definer.
For example, if a definition needs to be formulated for the term “journalist,” this existing day-to-day term must be explicated. Similarly, we must take into account various definitions that have already been created in the communication research. One first attempt to define a word could sound like this: “Journalists are persons who work for the media of radio, press, or Television.” Now we must verify if instances are possible that are covered by the definition but generally not considered to be journalists, and also if others are not covered but, nonetheless, are considered to be journalists. According to the present definition, a printer of a publishing house would be a journalist. An agency journalist would not be a journalist.
Then we must try an extended definition: “Journalists are persons who work for the media and collaborate with their fellow journalists on the editorial content of media products.” In this case, what is decisive is particularly the term “media product.” Even the simple and ordinary terms like “collaborate” are after all not unambiguous. The latter condition must be further refined: “. . . who draws at least one half of his income through his work for the media products . . .” This would solve other problems concerning the demarcations, but it could also pose new problems.
In this process, examples that generally do not belong to the term should be included in the same way as those that belong to the term. Whether an explication of a term covers all and only that which needs to be covered, remains always a hypothesis. Erroneous attributions cannot be ruled out with absolute certainty.
Real Definitions
By real definitions, the researcher attempts to determine the essence – the “true meaning” – of a term. There cannot be such real definitions, because no word has a meaning by itself. First of all, a link must be established between a word and its meaning. Often, explications of a term are hidden behind the supposedly real definitions, which – as explained earlier – are analyses of nominal definitions. Even if science does not have the sovereignty of definition over a term, the existing convention of meaning does not revert to something “real” but to the established usage of the term. These established conventions of language do not have, in terms of their validity, any higher precedence over the nominal definitions.
A similarly misleading argument for real definitions is the reference to acquisition of language prior to the age of comprehension among children. After all, a child’s first words are not based on any known terms. But yet there is an arbitrary definition that has no higher validity than the nominal definitions. To illustrate this: using a nominal definition, a “stool” can be defined as an “object that is about knee-high and has a seat but no backrest.” Or whenever a child sees a knee-high object that has a seat but no backrest, someone would say, “stool”. Also, every linguistic circumscription of a phenomenon is based on nouns and is, therefore, a nominal definition.
Another typical case of assumptive real definitions is the recourse to the history of the word; that is to say, the attempt to trace the meaning of the term from the etymology of a word. Apart from the change of meaning of words, an old definition is also based on an (obsolete) nominal definition, which, by no means, can claim a higher validity than the current one.
Operational Definitions
In the empirical research, terms must be determined so that they are measurable. We must decide how a characteristic should be precisely measured.
This process is called operationalization and the outcome “operational definition.” For example, if we are to find out from a man how long he reads newspapers daily, we can ask him: “How long do you read newspapers on average per day?”. The possible answers (“none at all”; 1–2 hours; 2 –3 hours . . . ) belong to the operational definition. If we wanted to know if someone is a journalist or not, we could ask him, for example, if he possesses a journalist’s identification. The characteristic “journalist” could be a trait if, for example, we are to ascertain whether journalists read more newspapers than all others. If I intended to conduct a journalist survey, I could use the same definition. So, we need operational definitions to determine the object of an investigation and to decide how the characteristics relevant to the investigation need to be gathered.
How do operational definitions differ from nominal definitions? Nominal definitions are expected to abridge elaborate descriptions through indicative terms, whereas operational definitions determine the measuring operations by means of which terms or constructs can be empirically obtained.
Unambiguity And Precision
A definition is good if it is unambiguous and precise. If, on the basis of a definition, we can decide whether a person, a group, an object, or an event is covered by the defined term or not, the definition is unambiguous. In principle, definitions always deal with a demarcation and, for that, necessary and adequate conditions need to be defined. The necessary conditions are meant for exclusion: any object that does not fulfill the necessary conditions of a definition does not belong to the defined term. If, on the other hand, the adequate conditions are met, the object is attributed, in any case, to the term. In this, the adequate conditions always cover the necessary conditions. Therefore, we must verify which conditions are applicable.
We have three options: (1) the necessary conditions are not applicable and the object will not be attributed to the term; (2) the adequate conditions are applicable (consequently, the necessary conditions are met) and the object will be attributed to the term; (3) the adequate conditions are not applicable, but the necessary conditions are: the object cannot be unambiguously attributed; it is placed in a gray area. The quality of a definition increases if the gray area between attribution and demarcation becomes narrower.
Validity
The operationalization is always measured on a parameter of ascertaining if it credibly covers what was meant by the nominal definition. This criterion of quality is called validity. If the operational definition defines measurement processes that unambiguously cover whatever is expected to be covered by the nominal definition, the operationalization is then valid. Validity is an assumption about the validity of a research operation. For example, if we were to ascertain if a person we are questioning is a journalist and, therefore, ask him if he possesses press identification, in the background the following assumption of validity is present: “every journalist obtains press identification and none other.”
If we intended to examine violence on TV, an operational definition could sound like this: “Whenever a person intentionally causes physical or emotional pain to another person, that is ‘violence’.” This definition may be adequate in many cases. Supposing we analyze TV news items where there is a report about fighter planes destroying bridges and roads. According to the present operationalization, no violence would be encoded here, but the group of researchers might agree without hesitation to extend the definition to “wanton destruction of property.” But then, we face a problem with medical drama on TV: doctors often cause physical pain to their patients in order to make a diagnosis or even during the therapy itself. The definition must be refined further. The validity of an operational definition can always be improved but it will always remain an assumption: exactly that and only that will be measured which needs to be measured.
References:
- Gerring, J. (2001). Social science methodology: A criterial framework. Cambridge: Cambridge University Press.
- Kerlinger, F. N., & Lee, H. B. (1999). Foundations of behavioral research. Belmont, CA: Wadsworth.
- Popper, K. R. (2002). The poverty of historicism. London: Routledge. (Original work published 1957.)
- Rubin, R. B., Palmgreen, P., & Sypher, H. E. (2004). Communication research measures: A sourcebook. Mahwah, NJ: Lawrence Erlbaum.