When the empirical material consists of a number of cases or objects - such as products, ways of using them, or customers - that are in some respects similar but in other respects not quite similar, you sometimes would like to describe the entire class by finding the typical case, from which all the other cases differ as little as possible. For much the same target, some researchers use the word "defining the essence" of a genre of objects.
In another project you perhaps are accepting that there may be important differences between the empirical cases, which difference usually becomes visible as a large variation in the measurements, too. In such a project you might want to point out, not one typical case but instead a few different typical cases, which approach sometimes is called constructing a typology for the material. It is discussed on the page Classification.
Either one of the above named targets of study is relatively easy to attain in the case that you can take as a support an earlier model or theory pointing out which characteristics of the cases are important and should be included in the description of the type. Other properties of the cases can then simply be disregarded in the analysis. - As a contrast, in exploratory study where no previous theory is available, it can be much more laborious to decide which data must be kept and which can be eliminated.
When selecting the method of analysis you must in any case pay attention to the nature of material that you have as your source. Usual alternatives in this respect are:
All the above alternatives are discussed separately below.
If at least some of the interesting properties of the cases have been measured, a quite usual procedure is to calculate an average for each variable, and then construct the "typical case" from these averages. For example, human measurements such as in the figure on the right, have already long been documented as averages for various populations (of different countries, of men and women etc). If the resulting type description seems to be lacking in detail, you can amplify it by copying non-numerical material from an empirical case that is near the averages.
In any case, it is advisable to have a look not only on averages but also on the dispersion of data: the less dispersion, the more credible will be description of the typical case with this method. If there is a large dispersion of data, you might consider the alternative of pointing out not one but several typical cases, i.e. a typology.
A usual method for finding out the typical shape among a number of objects is to superimpose the pictures like in the figure on the left. The original data in this example consisted of sketches made by a few inhabitants of Boston, showing how they perceived the structure of their city, in other words, what was their "mental map" of it.
In the resulting fuzzy model you can discern individual variation, but also those features where the respondents agreed. Exactly the latter was the invariant structure that the researcher was trying to define, and he finally draw it on an empty paper. (Steinitz, 1968.)
On the right is another example where several objects have been projected into one graphic. In the diagram, Sture Balgård shows how the old buildings in Härnösand follow uniform proportions of width and height (the red line) with just a few exceptions.
Other examples of pictorial presentation of typical objects and phenomena can be found later on in the paragraph Art as Analysis.
When empirical data consist not of measurements but of qualitative recordings, you have to construct the type description from those qualities and structures that are most frequent in the material. You can often start by coding and classifying the qualities that seem interesting, then notifying which of them are common to most cases and finally omitting those characteristics that are less frequent.
For presenting the combination of the typical characteristics, again, an artistic presentation could be worth considering, see Art as Analysis.
When studying objects that are common and well known to everybody, some researchers think that the best method for finding the essence of the object is not to gather material and restrict your study only to it. They think that, because the object is well known, you can get a generally valid view of it by taking as the object of study the "phenomenon" of experiencing the object, which you can approach by recalling all your earlier experiences of the object. The aim of this so called phenomenological approach, developed by Edmund Husserl during the years 1906-1936, is to obtain insights into the essential structures of these phenomena on the basis of mental examples supplied by experience or imagination and by a systematic variation of these examples in imagination. An advantage of this method is also said to be that it avoids the conceptual presuppositions made by earlier researchers which may be erroneous. This technique is certainly not always optimal for gathering empirical data, but its method of analysis has still some usability, besides being comfortable for the researcher who thus avoids awkward field work.
The basic method of phenomenological analysis is reduction. First, the existence of the object of study must be "put between brackets", not because the philosopher should doubt it but because the conditions around the object of study are subject to various coincidences which may obscure its real essence.
The second step is eidetic reduction. The aim is to find the universal and unchangeable essence or structure of the object. A suitable method is to imagine variations of the study object and focus one's attention on what remains unchanged in these variations.
For example, you might start by defining a 'table' as 'a horizontal sheet of wood with four legs'. Here, 'wood' and 'four' restrict too much and can be left out; in this way you will perhaps end at something like 'a horizontal sheet that stands on the floor' (as a contrast to 'shelf'). However, this method of eliminating the superfluous often produces quite a meager and not illustrative definition of the real thing. A remedy might be adding a picture or two as examples. If you have artistic talent, you might make a small work of art which combines details from several empirical cases. It would be no bad idea to include a few words about the degree of variation in the material.
Perhaps the best known example of the method is in the book Poetry, language, thought by Heidegger (p. 32, as translated by Albert Hofstadter):
"How shall we discover what a piece of equipment truly is? The procedure necessary at present must plainly avoid any attempts that again immediately entail the encroachments of the usual interpretations. We are most easily insured against this if we simply describe some equipment without any philosophical theory."
"We choose as an example a common sort of equipment -- a pair of peasant shoes. We do not even need to exhibit actual pieces of this sort of useful article to describe them." ... "A pictorial representation suffices. We shall choose a well known painting by Van Gogh." ...
"From the dark opening of the worn insides of the shoes the toilsome tread of the worker stares forth. In the stiffly rugged heaviness of the shoes there is the accumulated tenacity of her slow trudge through the far spreading and ever uniform furrows of the field swept by a raw wind. On the leather lie the dampness and richness of the soil. Under the soles slides the loneliness of the field path as evening falls. In the shoes vibrates the silent call of the earth, its quiet gift of the ripening grain and its unexplained self-refusal in the fallow desolation of the wintry field. This equipment is pervaded by uncomplaining anxiety as to the certainty of bread, the wordless joy of having once more withstood want, the trembling before the impending childbed and shivering at the surrounding menace of death." ...
"The peasant woman, on the other hand, simply wears [the shoes]." ... "When she takes off her shoes late in the evening, in deep but healthy fatigue, and reaches out for them again in the still dim dawn, or passes them by on the day of rest, she knows all this without noticing or reflecting."
Phenomenologists believe that their method gives deeper information about the empirical world than the usual empirical methods can give. Martin Heidegger (1972, 78) gave an "existentialist" explanation for this saying that a phenomenological insight comes from the existence (das Sein) common to both the researcher and the researched:
"Understanding is the same as the existential being related to a human being's own being and ability, and it takes place in such a way that this being reflects the meaning of its own existence back to itself. ... "When our understanding develops, we call it interpretation. In interpretation, we do not acquire knowledge on what we understand; instead, interpreting is all about realizing the potentials projected by understanding" (ibid., 82, 32).
In plain language, this seems to say that everybody finds it easiest to understand the products of his own culture, and the interpretation starts from the researcher rather than from the data. We can agree, but does the phenomenological method then give better information on the topic than other methods?
Heidegger does not clearly differentiate essence from the meaning of a word in a language. Although etymology seems to have been particularly appealing to Heidegger, the historical meanings of German words will not always tell speakers of other languages much about modern objects of study. When looking into the essence of concepts, phenomenologists are easily trapped by the etymological history of their mother tongue while there is no guarantee that the conceptual contents of just that particular language would be universally valid.
Another weakness of the phenomenological method is that almost any object seems to contain several essences. When for instance the researcher notices the phenomenon of children's red playhouse, he can easily reduce it in many alternative ways, for instance as:
Husserl's writings do not give a clear answer as to which alternative "eidos" should be chosen, because according to Husserl, the researcher should not take practical need as his starting point. Husserl thought that every research project necessarily involves one or another "research interest" which dictates which things shall be included. But one could ask why this research interest then need not be discarded?
In phenomenological study it is usual to try to improve the validity of the findings by contemplating the object in various contexts, notably in foreign spheres of culture or in ancient times. The argumentation behind this is the assumption that if different people during different periods have had similar conceptions of the object it shows a substantial invariance which can be accepted as a result of the study. Nevertheless, this logic includes the risk that the object has been understood differently in different times and cultures, in other words the object is not the same and there is no invariance after all.
The method of phenomenology is popular among dexterous essayists. When the same person asks the questions and gives the answers, all the pieces in the puzzle find their places. However, in practice, an alleged "phenomenological method" often just indicates that the writer wants to present the object in a new light and from a novel point of view and has made an extra effort to meditate its symbolical and associative attributes - often a laudable goal, of course.
It is obvious that even if the researcher, following the advice of Husserl, avoids empirical approaches in his study, his knowledge of the object is nevertheless always based on his earlier empirical observations. In other words, his knowledge is very subjective. But in phenomenological study, it is difficult for the researcher himself to see this because he must avoid criticizing the results he has obtained so that the sensitive and fragile insights would not elude him.
The aim of phenomenology seems to be the same as that of other research methods: eliminating details that are less relevant to the research project. If a phenomenologist succeeds in his aim and manages to define the essence or "meaning" of the object in what he deems as an accurate way, the reader should not be lured into thinking that all that is necessary has now been said about the issue. Many research objects are relevant to many people's -- not just the phenomenologist's -- lives in several important ways. One single researcher can seldom make a complete account of all these aspects.
The objective validity of phenomenological findings could be improved by discussing them in group. Besides, the very act of phenomenological discovery would probably be feasible in a modern innovative work group instead of the traditional lonely chamber of a philosopher, though this method has seldom been used. These two modes of work perhaps could be alternated. Another method to improve the validity of the results could be continuing the study with other empirical methods.
It is possible to search and present knowledge not only with the usual scientific methods, but also with some methods of arts. It is not usual today, but as a matter of fact art and science have common roots in antiquity, when the Greek term 'tekhne' and Latin 'ars' covered several areas of culture which only later were differentiated into arts and sciences.
Even today art and science have much in common, starting from their principal goals. The target of scientific research is to uncover and publish knowledge, information about the object of study, which knowledge then other people perhaps can use for solving their problems. This target is in principle similar to an important goal in art.
Another similarity is that the knowledge which is gathered and presented can be either descriptive, "disinterested", i.e. accepting the state of things as it is) or normative (i.e. explaining how you can change things).
As well in sciences as in arts we seek primarily new knowledge that has not been published earlier. Moreover, we consider that the better the presented knowledge is generalizable, the more valuable it is, because more people can then use it.
Some time ago most scientists and artists believed that there are two types of knowledge, one of them suitable to be presented as art and the other in the language of science. In works of art it has always been usual to present tacit knowledge that the artist has grasped from the surrounding world through empathy, or which he has inherited from earlier generations of artists. As a contrast, operating with tacit knowledge has been condemned in science. According to the so called positivistic school of thought it is the researcher's duty to explicate everything into plain language and disregard those things that he cannot explicate. "Whereof one cannot speak, thereof one must be silent" (Wittgenstein, B).
Today most scientists think that the positivistic ideal is feasible in some fields of study, but in other fields it is permissible to relax the requirements on explicity and precision of data if the alternative is to have no data at all. The attitudes towards tacit knowledge are thus today no more in sharp contrast in arts respective sciences. Lately, the Finnish university-level schools of art (University of Art and Design, Theatre Academy, Sibelius Academy and Academy of Fine Arts) have begun to accept even doctoral theses which consist of a "scientific" part and a parallel work of art which elucidates, exemplifies or complements the scientific findings. In this way, those research findings that can be explicated are placed into the "scientific" part and the remaining tacit knowledge in the artistic part.
Despite of the common goal of science and art - to present generalizable knowledge - their modes of presenting the intelligence are different.
A work of art presents information as a model of a singular case, but the mode of presentation is chosen so that it will be easy for the public to apply the knowledge to new contexts, for example to situations in their personal lives. One usual technique for this is that the artist first "studies" the motif on a more general level and then, when returning to the naturalistic level avoids unnecessary details in the presentation or makes deliberately the work ambiguous. It remains the task of the public, first to interpret the work of art into a more general level and thereafter to apply the content to their personal use (see figure on the left). These techniques differ from those used in the sciences, but the purpose is the same: make the model generalizable.
Scientific researchers, too, sometimes make individual case studies, the content of which is generalizable to some extent, but a more usual method for presenting general knowledge is to make a conceptual model, see diagram on the right. A usual technique to attain a generalizable scientific model is to detect invariances in data by removing random variation, 'noise' and 'disturbances'.
A conspicuous though superficial difference between art and science is in the mode of presentation of the knowledge. Each field of science, as well of art, possesses a long tradition for the mode and style of presentation. Engineers think that only measurements can give reliable information, while historians of culture believe that for catching the essential things you have to study qualities, not quantities. The mode of presentation marks also the boundaries between the various arts (like poetry, pictorial art, etc.) though today multimedia is tearing down this categorization.
In the dispute between artists and scientists it is easy to forget that all these languages present the same empirical world which in itself is neither "quantitative" nor "qualitative", "picturesque" or "profound". All these qualifications just refer to the styles of presentation used by artists and scientists. Usually the same information can be presented in several styles. You just have to choose which of them transmits best your message. As a matter of fact, most scientific types of general model (written language, icon, topological and analogous models) have originally been borrowed from various arts. The entire approach of case study is still very near artistic presentation.
Most scientists try to use as exact language as possible when presenting their findings. However, some researchers feel that a less exact presentation would better portray the subject matter. "Art can express not only ambiguity and ambivalence but also tension and contradiction - inevitable characteristics of our world and of the psyche" (Day Sclater, 2003, 623).
Traditionally we regard a skilful and impressive presentation as an essential property for a work of art, and the aesthetic pleasure that its perception gives to the public has been regarded as a measure of the quality of the work of art, cf. Beauty of a Product. It is interesting to note that an analogous "pleasure of discovering" can be experienced in sciences as well. A mathematical algorithm, for example the theorem of Pythagoras' triangle, can be perceived as 'elegant' or even 'beautiful' - one more similarity between art and science.
If your study includes not only descriptive but also normative approaches to art, you might be interested in the discussion its methods on the page Developing Art With Scientific Methods.
August 3, 2007.
Comments to the author:
Original location: http://www2.uiah.fi/projects/metodi