The Ragascape of Hindustani Music: III

A contextual perspective on ratings

Musicians select their repertoire for concerts based on certain contextual considerations.  It therefore makes sense to view the results of this study in that perspective. The classification attempts to follow contemporary practice (more than any theoretical construct), and is admittedly debatable.

 

The rating scores reported here are aggregates for each of the groups. The classified ratings of the 97 Raga-s are reported in Table 3.

The early morning group shows a substantially lower instrumental music score than the vocal music score. This is interesting, as early morning Raga-s are mostly associated with the devotional sentiment. The human voice may understandably be perceived in the culture as a more appropriate expression of this sentiment than an acoustic machine.

The instrumental/vocal music ratio rises steadily after the early morning segment to reach its peak with the late night group of Raga-s. The pattern is interesting, though not sufficiently strong to be considered a phenomenon.

The Thumree group, shown to lead this table, is so called because the Raga-s in that group are encountered largely, though not exclusively, in the romanticist genres – Thumree, Dadra, Hori, Tappa etc. Membership of this group includes Raga-s like Bhairavi, Khamaj, Kafi, Gara, Piloo, Pahadi. These Raga-s are versatile because of their malleable grammar and melodic lyricism. This group is rated generously because it covers three components – actual Thumrees performed by vocalists, instrumental renditions of Thumrees and allied genres, and classicist presentation of “Thumree” Raga-s by vocalists and instrumentalists.  The emergence of this group at the top may suggest, that the present-day musical culture relates more readily to the vivacious flavors of Hindustani music than to its other moods.

Raga groups at the bottom of the ranking – Early morning and Sunset Raga-s – appear to be telling the same story, though in a different way. Early morning Raga-s are mainly of the Bhairav and Todi families, while the Sunset Raga-s are mainly of the Puriya/ Purvi group. Both periods are earmarked by tradition for religious/ devotional activity, and the prescribed Raga-s possess melodic features suggestive of piety. Our study pushes them to the bottom to suggest that the present-day musical culture has little interest in solemnity.

The moderate-to-high rating in the late evening and night groups are as expected. A vast majority of concerts are held during these periods, thus making for a wider spectrum of familiarity amongst audiences. The choice of Raga-s for performance during these periods is also vast, thus creating an equitable distribution of audience involvement within the category.

Predictably, the comfort level with Carnatic Raga-s in their Hindustani manifestations is uneven. The classification of Raga-s within this group relied on the authority of Raga Nidhi (B. Subbarao, Music Academy, Madras). Hansadhwani (1782 rating points), Keerwani (874 rating points) and Charukeshi (448 rating points) appear to have, by now, evolved a convincing interpretation in Hindustani music. Others, imported more recently, may either take longer to achieve this, or fade away.  

... Continued in Part IV

The Ragascape of Hindustani Music: II

Patterns in Audience Engagement

Table 1 presents the list of the 97 considered Raga-s in descending order of views per month (the audience engagement indicator). The indicators were also computed separately for vocal music and instrumental/ orchestral music.

The composite Audience Engagement Indicator across 97 Raga-s is 526 views per month, incorporating a vocal music score of 462 and an instrumental/orchestral music score of 645 views per month. The composite scores range from 13 views per month (Raga Jaitashree) at the lower end to 3475 views per month (Raga Bhairavi) at the upper end.

The higher aggregate rating for instrumental music can be misleading because the figures are distorted by a few extreme outliers related to events abroad.  A more realistic picture is obtained through a correlation run on the two series. It is clear that vocal music supports the composite rating of individual Raga-s more categorically than instrumental music. (The vocal music rating has a 0.94 coefficient of correlation with the composite rating, while the instrumental music rating has a correlation of 0.85.)

Composite Rating of Raga-s

 Audience Engagement Indicators for the 97 Raga-s yield a median value of 272 views per month, against an aggregate score of 526. Graph 1 displays the frequency distribution of the 97 scores, presented at intervals of 200 points.

 This is a highly skewed distribution. Interpreting such distributions can get quite complicated.  To keep things simple, we may use popular measures such as mean (490) and standard deviation (608). One simplistic (and scientifically questionable) interpretation would be a cut-off point at 1098 (Mean + Standard Deviation). This cut-off point would leave us with 11 “Prasiddha” Raga-s out of 97 Raga-s. A different, but “common sense”, view can permit a cut-off point at the mean (490). This point would give us with 28 Raga-s with “above-average” audience engagement indicators, and bring us closer to the Bhatkhande ratio.  

This “cut-off point” approach is, of course, unfaithful to the reality. In the real world, Raga-s have no appeal independently of the musician. Musicians shape Raga-s as much as Raga-s shape music. Time and again, it is proven that the comfort zone boundaries of audiences cannot resist the power of musicianship. Why, then, do we mention such a simplistic notion? We do so because musicians find its simplicity more appealing than the calibrated approaches of academics, and because it does not weaken our basic argument.  

                                                                                         

This “above average” 28-Raga “Blue Chip” list would also fit neatly into a conventional analytical concept – “Share of views” (unadjusted for the duration since upload). Table 2 compares the duration adjusted views (views per month) for each Raga with the percentage share of gross views as accumulated up to the date of audit for the top 28 Raga-s reporting an above-average audience engagement indicator. The Table shows the two columns of ranking, representing two different approaches to the measurement of audience engagement, running almost parallel in terms of relative values. The cumulative share column in Table 2 suggests that these 28 Raga-s together account for over 70% of all views logged for the 97 Raga-s.

These results can be broadly interpreted as saying that 30% of the Raga-s account for 70% of audience engagement, while the remaining 70% of the Raga-s collectively account for only 30% of audience engagement.

The five Raga-s at the top of the rating have been widely heard in popular, semi-classical, and film music for as long as any living person can remember. This is a useful reminder that Raga-s are not the exclusive property of classical music. The dividing line we often draw between different categories of music (classical/ popular/ devotional/ martial etc.) is an academic construct. Raga-s reside in the racial memory, and are accessible to anyone seeking them. The Ragascape is shaped by all categories of music and, in turn, shapes all categories of music, though in different ways.

On close scrutiny, the 28 Raga-s in Table 2 reads like a sensible prescription of “standard repertoire” for a professional musician. The selection is well distributed across the various segments – Early morning (2), Late morning/ afternoon (7), Sunset group (2), Late evening group (7), Night group (2),  Thumree Raga-s group (4), Seasonal group (2) and Carnatic group (2).

The Raga-s at the bottom of the heap (Table 1. Rank 77 through 97) also reveal a fairly even distribution across categories. Early morning (4), Late morning/ afternoon (3), Sunset Raga-s (2), Late evening (4), Night Raga-s (1), Thumree Raga-s (2), Seasonal Raga-s (2) and Carnatic Raga-s (2).

It is natural for any curious mind to ask – what factors determine the rank each of these 97 Raga-s hold in the present output? The study is not designed to answer this question. What we have here is an audience engagement indicator as derived at the time of the study.  The resultant ranking is useful today and, perhaps in the immediate future. We know it is volatile; but we do not know how volatile.  

A Shrinking Ragascape?

These results tend to confirm the belief that the contemporary Ragascape is fairly narrow, with perhaps just about 50 Raga-s accounting for almost all performances, across all media.  It also supports the suspicion that enviable careers can be built relying on a repertoire of 12/15 Raga-s.  

One can argue that a narrow Ragascape enables musicians to aim for progressively greater depth in the exploration of a few Raga-s, in preference to achieving only a superficial treatment of a wider repertoire. This sounds reasonable if we are referring to a single musician’s career strategy/ choice. But, the argument loses traction if it tries defending a collective phenomenon – such as we have identified here. If audiences are being fed constantly on re-packaged doses of the familiar, they will tend to become indifferent – if not actually averse – to novelty/ variety. When this happens, musicians lose interest in enlarging their artistic resources. The natural consequence is the atrophy of imaginative capabilities. An imagination deficit is the surest path to artistic sterility. More fundamentally, then, who needs classical music?

We cannot ascertain whether the Ragascapes of the past were wider than they are today, or narrower. On reasonable reckoning, a shrinkage began in the 1960s, when All India Radio began to withdraw from its role as the dominant purveyor of   Hindustani music, leaving the recording companies free to fill the vacuum. The Long Playing record and subsequent storage media innovations – providing concert length recordings -- made this usurpation easy. Volume-driven strategies of the recording companies lead them to concentrate their resources on star musicians. Fewer musicians on the market meant fewer Raga-s in circulation. As a result, five decades later, we have two generations of audiences (and also musicians?) whose comfort zone may not exceed 30-40 Raga-s.

YouTube and other online repositories may alter this picture in the future because they appear to attract a growing diversity of content to achieve their commercial goals. But, we cannot be sure.  The evolution of these platforms is being guided by sophisticated Machine Learning applications designed for maximizing profitability globally. Even the designers of these AI systems cannot foresee how their systems will shape music across cultural boundaries.

The Top 28 ranked Raga-s.

... Continued in Part III

The Ragascape of Hindustani Music: I

About a hundred years ago, VN Bhatkhande published compositions in 181 Raga-s, of which he described 45 as being “Prasiddha”. The remaining were classified as “Aprasiddha”. This classification is often mentioned in contemporary musicological discourse, along with an unstated question: If a similar classification were to be attempted today, what kind of numbers would it yield? This question inspires this enquiry.

Bhatkhande was a Sanskrit scholar. What precisely did he mean? He did not use “Prakhyat” (प्रख्यात) = famous. He did not use “Prachalit” (प्रचलित) = common/ popular. He did not use “Pratishthit” (प्रतिष्ठित)= respected/ elitist. He used “Prasiddha” (प्रसिद्ध), derived from the Sanskrit “Siddha” (सिद्ध्) = established/ proven. “Prasiddha” carries all the connotations of the candidate words not used, but elevates the issue to a different level. The “Siddha” quality here has two angles – of mastery by performers, and of being proven to listeners. It denotes fame attained through proven excellence/ mastery, and wide acceptance/ endorsement by public opinion. In the present context, therefore, Bhatkhande’s notion of “Prasiddha Raga-s” may be interpreted as Raga-s established in the musical culture on account of accomplishments (of musicians, in performance), and their widespread acceptance by listeners.

To arrive at his assessment, Bhatkhande interacted extensively and intensively with musicians and other influentials, such as scholars and patrons, across the country for several decades. His was the most authoritative assessment that could have been made of his times. But, by today’s standards, the music world he surveyed was small, and uncomplicated. Today’s Hindustani music is a vibrant sub-culture -- larger, more complex, more diverse, and connected by a more transparent web of self-interest amongst its participants. It neither asks simple questions, nor accepts simplistic answers.

Today’s professional musician is a service-provider in the market for Raga-based music. He has a well-developed notion of “career strategy”. The strategy accounts for the repertoire in which he has a high level of performing competence. But his “command repertoire” does not necessarily ensure professional success.  His success also depends on the receptivity of the contemporary Ragascape – the set of Raga-s within the “comfort zone” of contemporary audiences.

The idea of a Ragascape connects with the notion that Raga-s are archetypal entities residing in the racial memory as distinct banks of musical ideas, each associated with a particular region of the community’s emotional life. Musicians draw upon these banks to relate meaningfully to their audiences and, while doing so, also contribute new ideas to these banks. Raga-s drift in and out of circulation and public consciousness through community-wide recalling and forgetting. Conceptually, then, the Ragascape of an era is comprised of Raga-s whose ideational banks are being actively/ frequently/ constantly tapped and replenished in the present.

Veteran observers of the music scene have noted that, from a universe of possibly 1000+ documented Raga-s in Hindustani music, not more than a 100 are “in circulation” at any given time – say, a generation understood as 30 years. There is also considerable agreement that, of these 100, perhaps just about 50 Raga-s account for the vast majority of performances across all media. If informed-jury estimates are to be believed, 50 to 100 Raga-s may be sufficient to paint the Ragascape of any era.  

Whatever the truth, it has to connect with the reality of the musician being an economic being.  Throughout his career, he explores and cultivates a “goodness of fit” between his artistic resources and public receptivity through some awareness of contemporary trends. If he has the makings of an epochal musician, say of an Ameer Khan, Alladiya Khan, or a Faiyyaz Khan, he may actually alter the Ragascape. But, even these worthies had to reckon with the Ragascape of the era, before they began to alter it. They obviously did some good, even if informal, “market research”. Their kind of research can now be a little more “formal”.

The Ragascape idea also impinges upon the grooming of aspirants for careers in classical music. Individual Gurus and educational institutions impart skills either without utilitarian considerations, or without organized knowledge of the Ragascape in which their wards will seek a career. With even imprecise information, educators can enhance the value of their training. It is, of course, true that the worldview of the active Guru-s (and teaching institutions) in each generation will collectively shape the Ragascape through what is taught and how. But, this is a two-way process. For every musician, the Ragascape is a “given” reality, which also permits him to enrich or alter it.  

The dynamics of the Ragascape are surely of interest to musicologists. Raga-s drift into, and out of, the Ragascape in response to changes in aesthetic values, and the changing profiles of musicians and audiences. Every Raga has its distinctive personality and performing stance. Every Raga entering or exiting the Ragascape says something about society and its expectations from classical music. Such clues are mostly encountered without being sought, and deserve more than a glance from scholars.

The challenge, then, is to devise a methodology for mapping the Ragascape to provide useful insights to the various participants in the Hindustani music ecosystem.

The contemporary Ragascape is the cumulative product of Raga-s performed with distinction by several generations of musicians working with a multiplicity of Raga-based, (and even Raga-neutral) genres, and delivered to audiences through a variety of personal and impersonal media. Mapping such an activity involves musicians and audiences scattered all over the world. This is a formidable task for which neither the methodology, nor the funding, can be envisaged. One may, therefore, consider existing data sources that can yield workable, though not authoritative, insights.

This study draws on public-domain information on YouTube viewership to extract some insights that may be helpful to the classical music community. YouTube data is increasingly being used for musicological research, while its value remains a subject of academic debate. There are good reasons to utilize the data prudently, and with an awareness of its limitations. Appropriately, then, the numerical outputs are to be interpreted as “Orders of magnitude” rather than categorical measures of what is being specifically measured. The limitations of using YouTube data have been discussed in detail at the end of this study.

The data used here for computing the Audience Engagement index of Raga-s was capable of yielding a few additional insights. Such leads have been pursued to the extent the data would permit.

YouTube as data source

YouTube is, by now, the largest open-access repository of Hindustani music, hosting recordings going back to the earliest years of sound recording. It appears to have replaced every other access to Hindustani music – other than, possibly, a live concert. But, increasingly, any significant (or even insignificant) contemporary concert or recording taking place in any part of the Hindustani music world, finds its way to YouTube before long. YouTube may now be considered more than merely “representative” of contemporary Hindustani music – it could possibly qualify as the primary platform. The long-term/ cultural implications of this phenomenon need not concern us here. It suffices to acknowledge that YouTube audience measurement numbers can be treated with respect.

Could the flood of live Hindustani concerts on Facebook during the global pandemic shut-downs (April-May-June 2020) have diluted YouTube’s status as the dominant music platform? There is enough reason to believe that, fundamentally, nothing has either changed or is likely to change. However, it is safest to state that we do not know; or that it is too early to judge. In any event, this present study is insulated against even a temporary disturbance on this account. The research database was compiled between March 25 and April 5, 2020, before Facebook became an online auditorium.

YouTube remains a viable business by selling advertising exposure. In order to maximize advertising value, it operates a sophisticated system of managing viewer behavior. The system uses internally generated information and navigation prompts to guide users into spending the maximum possible time on YouTube. The same system provides advertisers with reach and impact data on the content to establish its advertising value. Except perhaps under special arrangements, the analytics generated by YouTube are not available to researchers outside its circle of commercially valuable users.

The data available publicly on screen is all we have. We have (i) the date of upload, and we have (ii) the total viewership from the upload date till the day we are looking at the numbers. This study draws on these fragments of information for pursuing its query.  

 The Audience Engagement Indicator

The Raga-s considered for rating for audience engagement were selected by merging the under-graduate and post-graduate syllabi of the major universities and examining institutions involved in Hindustani music. This framework was valuable as an indicator of organized activity aimed at preserving their character, and keeping them “in circulation”.

The computation of the Audience Engagement Indicator is based on the belief that Hindustani music is a highly individualistic art, and hence audience loyalties/ preferences are centered upon individual musicians rather than around Raga-s. Isolating Raga-specific engagement from viewership measures needs mountains of data and some arithmetic. The objective is substantially achieved by aggregating the metrics of videos of the same Raga performed by a large number of musicians across various genres of music, performing in various contexts, across diverse geographies. A total isolation of the Raga effect from the musician effect is neither possible, nor necessary. Musicians shape Raga-s, as much as Raga-s shape music.

The arithmetic is simple. The videos covered for this study were uploaded at different time distances from the date on which we are logging their total viewership. Each video has thus had a different time-span over which to accumulate viewers – or be forgotten. So, the aggregate viewership of all included videos has to be adjusted for these differences in order to obtain a standardized measure of audience engagement. The resultant number is computed as “Views per Month”. However, it is safest to regard it as an unrefined “Audience Engagement Indicator”.

  It is fair to ask whether what we are measuring is, in fact, what we wish to measure. The truth is that we neither know, nor can alter, the manner in which YouTube measures and reports viewership on the screen.  We are only incidental beneficiaries of the information available, and of whatever meaning we can extract out of it. The limitations of the data source are discussed in some detail at the end of this study. 

... Continued in Part II