134 Agnes Petocz
There are fairly strong theoretical, philosophical and pragmatic rationales for examining
the concept of objectivity and finding it wanting ... But validity is a more irritating
construct, one [not] easily dismissed ... [because it is concerned with the question] are
these findings sufficiently authentic ... that I may trust myself in acting on their
implications‖. (p. 178)This is clearly what led W. Potter (1996) to be "bothered by the seeming lack of
correspondence between what qualitative theoreticians prescribed and what practicing
qualitative researchers did" (p. 234). He notes that "researchers must realize that some of the
prescriptions cannot be followed" (p. 327).
The proponents of qualitative methods do not consider realism as the way out of the
"crisis" or "dilemma". Consequently, their proposed solution, in the form of pragmatism as
the dominant paradigm for the ―mixed methods‖ research promoted by the ―third
methodological movement‖ (Tashakkori & Teddlie, 2003) simply exacerbates the situation by
perpetuating the same confusions. Consider some of the supposed benefits of adopting
pragmatism to implement the "fundamental principles" of a mixed methods approach. First,
we are told, ―Pragmatism avoids the use of metaphysical concepts (e.g., ―truth‖, ―reality‖)
that have caused much endless (and often useless) discussion and debate‖ (Teddlie &
Tashakkori, 2003, p. 21). Next, we are told, one of the fundamental principles is that ―Data
collection procedures are independent of data analysis techniques ... Collected data may be
transformed at any point ... and may be analysed both quantitatively and qualitatively‖
(Tashakkori & Teddlie, 2003, p. 696, emphasis added). In other words, one can quantitize
qualitative data and qualitize quantitative data (Teddlie & Tashakkori, 2003, p. 9) – in
accordance with the thesis that ―the difference between quality and quantity is not a
difference in nature, but a difference in our conceptual schema – in our language‖ (Carnap,
1966, p. 58). This antirealist conventionalism is then contradicted in the very next
―fundamental principle‖ of mixed methods research, which is that: ―If the data do not
represent the theoretical phenomena or the attribute under study, then nothing else in the
design of the study matters‖ (Tashakkori & Teddlie, 2003, p. 696, emphasis added). For,
under the Carnapian conventionalism and ―principle of tolerance‖, the data must always
succeed in representing the phenomena, for they do so by fiat.^22 The possibility of their not
doing so is allowed only by realism; only a realist approach locates quality and quantity as
real features in the structure of real situations, such that we cannot simply change a quality
into a quantity (or vice versa) by what we do with numbers or with words by way of
representing those features.
Of course, the real problem here is that proponents of qualitative research do not realise
that a thoroughgoing realism, with its conception of reality as consisting of complex and
contextualised situations, and its broader conception of science, is perfectly well suited to
acknowledging and dealing with layers of meaning, multiple interpretations, social
constructions, subjective perspectives, data uncertainty, cognitive fallibility, etc. As long as
the conception of science and realism are appropriately broad and flexible, there is nothing at
all to be lost by adopting them for qualitative research, and everything to be gained - for the
result is dissolution of the tension between what is practised (method) and what is preached
(philosophy). Therefore, the "methodologically aware eclecticism" argued for by
(^22) Just as the variables ―measured‖ in quantitative psychology are rendered "quantitative" by the measurement
operation.