on an organizational level
4.8 Validity, reliability, and generalization
The three terms validity, reliability and generalization have turned into what Kvale calls the holy trinity of science.157 Qualitative studies often involve
case studies looking at a single environment. To claim that your findings from a local, limited environment can have global value is difficult, and often critics of qualitative research can be found making this challenge. How can my findings, from my studies of a few vocal ensembles with a focus on one of them, have some kind of global value? The easiest thing would be to deny that the generalization of research findings is a goal at all, and there are researchers who try to minimize the relevance of generalization.158
Payne and Williams (2005) admit that generalization remains a big problem in qualitative research,159 but nevertheless they suggest a moderate
generalization through, among other suggestions, deciding upon how
widely applicable the findings should be, limiting the time period of the findings, limiting claims to basic patterns, or saying that the nature of the generalization will be conditional upon the ontological status of the phenomena in question.160 In this present study I have mainly followed
nvs. But I have also followed other ensembles – I Fagiolini and The Hilliard Ensemble. With the addition of Nordic Voices, my own ensemble, these ensembles form quite a good representation of the classical vocal ensembles of the world. I look at various aspects of vocal ensemble practice from a theoretical and practical point of view, as Payne and Williams suggest, moderate generalization is possible through, among other things, a firm theoretical foundation. Therefore, I will not completely resist generalization, but suggest moderate generalization where I consider it appropriate.
157 Ibid. p. 160.
158 Payne and Williams 2005 p. 295. 159 Ibid. p. 309.
Frank Havrøy: Alone Together
In daily speech, when an argument is valid, it is considered as the truth, or at least something trustworthy. A valid argument is a reasonable, strong and convincing argument. Kvale says that to validate something is to control it.161
It is the researcher interpreting his own findings with a critical mind. Kvale also says that to validate is to ask questions.162 For instance, an interview
subject might not tell the truth, and this has to be controlled each time. Kvale argues that to validate something is to theorize it,163 meaning that the theory
chosen for a study is connected to the method. Kvale differentiates between communicative validity (meeting an argument with the opposite argument), and pragmatic validity (that action follows knowledge).164
As Merriam says, internal validity asks the question: how congruent are one’s findings with reality?165 Merriam suggests different approaches to
strengthen the internal validity of your findings, using triangulation (the use of multiple investigators, multiple sources of data or multiple methods),
member checks (taking the data and the findings back to the participants,
asking if the interpretations are plausible), peer/colleague examination (asking peers or colleagues to examine the data), statement of researcher’s
experiences, assumptions, biases etc. (enables the reader to understand how
the data was interpreted), and submersion/engagement in the research
situation (collecting data over a long enough period of time to ensure an
in-depth understanding of the phenomenon).166 In this present study I
have employed a multiple method angle in the collecting of the data. I have studied vocal ensemble practice and nvs in interviews, observations, video analysis, and more experimental methods. This should cover the demands required for thorough triangulation of the material.
‘Reliability is concerned with the question of the extent to which one’s findings will be found again’.167 This is a direct quotation from Sharan B.
Merriam (1995), and it means that the findings of a study should be the same, if the inquiry were replicated. This is a challenging criterion, as qualitative research always has to do with human beings and human phenomena. If
161 Kvale 1997 p. 168. 162 Ibid. p. 169. 163 Ibid. p. 170. 164 Ibid. pp. 170–176. 165 Merriam 1995 p. 53. 166 Ibid. pp. 54–55. 167 Ibid. p. 55.
my research work in this study was replicated, the results could very well be a different interpretation of the findings, but still true. Instead of using the term reliability, Merriam suggests using ‘dependability’ or ‘consistency’, meaning that the results of a study are consistent with the data collected.168
In this context, triangulation is central (see above).
4.9
Conclusions
In this chapter I have shown the methodical framework employed in this study, focused on a case study of nvs. Within this methodology there is a multiple approach, including interviewing of participants, observation, participant observation, video analysis, and also elements of methods like stimulated recall and more experimental strategies. Using these methods, I have chosen different approaches to the analysis process, and through this conglomerate of methods, my findings are triangulated, making the study both valid and reliable.
Whereas others might cruise along on a diet of antique favourites and the occasional foray into the twentieth century, the Courage Consort were always open to a challenge from the avant-garde. (Faber, 2002)
I
n one of my first conversations with Christine Fischer, the administrative leader of nvs, she talked about how the classical vocal ensemble had gained quite a new position in musical life during recent decades. Since 1950, she claimed, composers had started to write for these ensembles again, and the repertoire for these ensembles had therefore slowly grown. This had not happened, according to her, since the Renaissance, when composers wrote for constellations containing one singer per written voice. I immediately recognized this attitude. It confirmed my own experience about the vocal ensemble’s history. In Nordic Voices we had long ago defined our activities as taking place within two fields: early music, meaning music from around the thirteenth century and until around 1750 and contemporary music, running from 1950 up until today. Even if we could from time to time throw in a piece by Josef Rheinberger or Johannes Brahms, it was always with a feeling that we were doing something that Nordic Voices was not created to do, something that did not suit the ensemble. In discussions around this repertoire, two issues were often raised: the repertoire was not originally written for this kind of ensemble, but for a bigger choir, and therefore the sound would be significantly different from the sound the composer had intended. And, when it came to Romantic music, there was a feeling that our ‘thin’ vocal ideals, light on vibrato, would be a long way from the vocal ideals traditionally expected for the music from the Romantic era. We felt as if we represented a complete different kind of vocal tradition. A searchFrank Havrøy: Alone Together
through the vocal ensembles I knew about throughout the world told me that almost every one of them dedicated themselves to either early music or contemporary music or both categories. It confirmed my assumption: vocal ensemble singing, meaning classical vocal music sung with one voice per part, seemed to be best suited for singers who had chosen another path vocally than the traditionally taught classical solo singing style, which to our view was concentrated mostly on music between 1750 and 1920.
The members of nvs also followed this path. One of the singers, when questioned about why the vocal ensembles of the world do not sing music from between 1750 and 1920, answered that the pieces from that time were meant for larger ensembles, that they were written for ensembles with more people than one voice per part. Another singer stated that there was a shift of focus through the baroque era, from smaller groups to larger and larger groups, from vocal ensembles to choirs. You could find composers who wrote for solo quartets, terzettos and so on through the Romantic era, but preferably accompanied by piano. Again the singers of nvs would underline that they represented a different singing ideal to that of traditional classical singers.
When asked about concrete elements in their vocal technique that differentiated them from other traditional, classical solo singers, they all pointed to a singing ideal that was ‘more flexible, slimmer and with a definite control over vibrato’.
There seemed to be consensus among the vocal ensembles of the world that the repertoire of this period was not intended for the ensembles, and that the singing styles from this period were quite far from the vocal aesthetics held by those in vocal ensemble practice. From this I concluded that what I called the Romantic era was not the era of the vocal ensembles, and that there had to be quite few, if any vocal ensembles operating during these years.
But then I stumbled across the Halfdan Kjerulf Quartet in Norway. This ensemble, with up to eight singers, operated in Norway during the 1850– 1860s. At the same time I discovered that the famous singer Lauritz Melchior had been a part of a quartet called the Scandinavian Quartet, together with the singers Olaf Peelmann, Holger Madsen and Holger Hansen. They had toured Sweden in the early years of the twentieth century, and had even
recorded some of their music.169 Both these ensembles proved that classical
vocal ensembles had existed at least in the last half of the nineteenth century. Both ensembles seem to have focused on the Romantic male choir repertoire, and they proved two things: that singers with ‘Romantic’ singing ideals absolutely could sing in a classical vocal ensemble and that professional vocal ensembles were present in a period from which many vocal ensembles of today have distanced themselves.
The existence of these ensembles could be an indication of a vocal ensemble practice also present during the ‘Romantic’ era, quite in contrast to my own assumptions about vocal music in this period. The question I felt arose was that of why the vocal ensembles of today distance themselves from the repertoire of the period between 1750 and 1920.
I have chosen to have a small look at the history of the vocal ensemble practice, with attention to what I call the ‘big gap’ in the repertoire of the vocal ensembles of today. This historical view will only serve as an illustration to the questions being asked about this ‘big gap’. I am not telling the whole story about the vocal ensemble practice from the Middle Ages to the present day, since such a survey would exceed the limitations of this thesis. But I will underline that my view is that the vocal ensemble practice seems to not have been a prioritized area for music historians, and I hope that there will come more historically founded projects on vocal ensemble practice in the future.