Thursday, February 18, 2021

The Ragascape of Hindustani Music: A Composite View (Concluded)


In this concluding part of the Ragascape studies, I attempt to integrate the results of the three Ragascape studies reported earlier, into a composite perspective on the phenomenon being explored.


In the first study, we identified 97 Raga-s prescribed for study at under-graduate and graduate levels by students at the major universities granting degrees in music. For this set of 97 Raga-s, we computed “Audience Engagement Indicators”, drawing on YouTube viewership data irrespective of artist profile.


In the second study, we audited YouTube archives to identify the Raga-s that have been performed at eight of the major music festivals of Hindustani music over the last 15/20 years, and to rank them in order of frequency of occurrence. This study did not restrict itself either to the Raga-s considered, or to any specific profile of the musicians performing them.


The approach to the third study was guided by the results of the first study. The first study had found that, contemporary musicians achieve a much higher “Audience Engagement Indicator” on YouTube than earlier generations of musicians. Responding to this trigger, the third study identified 20 leading contemporary vocalists and 20 leading contemporary instrumentalists, and used YouTube archives to enumerate how many of the 40 selected musicians have performed which Raga-s, irrespective of context.


These three studies gave us three different rankings of Raga-s, each obtained from a different perspective on the Ragascape phenomenon. The task before us now is to integrate the three different perspectives into a composite measure of each Raga’s presence on the Ragascape of Hindustani music.


Integration of perspectives


The three studies gave us three rankings, each from a different perspectives. It was found that 81 of the original 97 “candidate” Raga-s appeared in each of the three lists, though the rank of each Raga in terms of importance could differ. Computing a composite measure was, therefore, possible only for these 81 Raga-s.


In order to integrate the three rankings, they needed to be brought to a common mathematical platform. The simplest method for doing was to re-calibrate each of three measures/ rankings with the highest as Base=100. The three indices so derived were averaged in order to derive the ‘Composite Index”. The three individual indices for 81 Raga-s, along with their “Composite Index” rating are shown in Tables 1A and 1B. The Coefficient of Correlation between the three component indices within themselves, and with the Composite Index are given in Table 1C.


The correlation patterns suggest that (1) the individual indices are reasonably confirmatory of each other, and yet distinct in terms of reflecting the different approaches to viewing the Ragascape phenomenon and (2) the Composite Index is equally reflective of each component index considered separately.


Patterns in the Composite Index


The purpose of computing the Composite Index is to prioritize/ rank Raga-s in terms of their dominance on the Ragascape, and identify that group of Raga-s whose presence contributes disproportionately to shaping the musical culture. For this purpose, the distribution of the Composite Index was considered through the Mean, Median, Standard Deviation measures, and also the contribution of each Raga to the Aggregate Composite Index. The analysis is presented in Tables 2A, 2B, and 2C.


After considering the three different approaches, it was found that the median value of the composite index gave us the sharpest differentiation between the Raga-s that dominated the Ragascape, and those that trailed the list of 81 Raga-s. A total of 42 Raga-s are found to have index values equal to, or higher than, the median value. These 42 Raga-s may be considered to constitute the core of the Ragascape, and as reflecting the contemporary musical culture. The analysis of these 42 Raga-s provides some insights into the melodic canvas of contemporary Raga music. The various ways of looking at this phenomenon are considered in Tables 3A, 3B, and 3C.


At this stage, it must be clarified that the bifurcation of the 81 candidate Raga-s was not guided by any desire to confirm, or even approximate, Bhatkhande’s listing of 51 “Prasiddha” Raga-s. It was guided entirely by the mathematical logic of the composite index as given in Tables 3A, 3B, and 3C. The “cut-off” point had to be placed at a point below which there was a sharp drop in the average composite index.


Analysis by broad categories (Table 3A)


The rating of Raga-s on the Composite Index by broad categories is dominated by three categories -- Thumree Raga-s, followed by late morning/ Pre-sunset Raga-s, Carnatic Raga-s, and late evening Raga-s. The domineering presence of Thumree Raga-s (Bhairavi/ Khamaj/ Kafi/ Piloo etc.) suggests a musical culture tilting categorically towards the romanticist genres and relatively undemanding music – undemanding from the listener’s perspective. The significant presence of late morning/ pre-sunset Raga-s, and late evening Raga-s reflects the preponderance of concert events during these hours of the day/ night. The poor presence of early morning and sunset period Raga-s confirms this phenomenon.


The low rating of seasonal Raga-s is to be expected, since they are performed only during the appropriate seasons. But, the strong presence of Carnatic Raga-s is interesting.  Hansadhwani ranks at no.9 and Kirwani at no. 21, on the composite index, and a substantial number of leading contemporary musicians do perform them. This is a good reflection of the success of Hindustani music in imparting to a few Carnatic Raga-s a mature melodic personality during the post-independence era.


Analysis by melodic structure (Table 3B)


Amongst the various classifications of Raga-s in the Hindustani tradition, an important classification is based on the number of swara-s deployed by a Raga in the Aroha (ascending melodic motion) and the number deployed in the Avaroha (descending melodic motion). This dimension is important because it defines the improvisational potential of a Raga, and its melodic complexity. Any Raga can be either Audav (pentatonic) or Shadava (hexatonic) or Sampoorna (heptatonic) in one of the two directions or in both directions.  This facet of Raga grammar has a substantial bearing on its melodic character.


The two categories dominating the top-42 list are interesting in this respect. Both are identical in ascending and descending melodic motions. The relatively low-rated categories exhibit an asymmetry between ascents and descents.


The leading category, on composite index rating, are the Sampoorna-Sampoorna melodic entities. Being heptatonic in both directions, these Raga-s have vast potential for improvisation. Though these Raga-s are perhaps more demanding in terms of melodic ideation than other categories, their grammar is less demanding than Raga-s, which are asymmetrical in their directional patterns.


The second largest score on the Composite Index is enjoyed by the Audava-Audava category – pentatonic in the ascent and pentatonic in the descent. Featuring only 5 swara-s (the minimum required for a Raga), these Raga-s represent, by definition, the simplest melodic structure for musicians to handle and for audiences to comprehend.


Analysis by tone-material typologies (Table 3C)


Modern (20th century) music theory classifies Raga-s into different typologies based on the swara-s deployed in the conception of each Raga. A macro-level classification, propounded by Bhatkhande, was the system of 10 “Thaat-s”, which is controversial. Around the same time, there also emerged a micro-level classification, which is associated with the convention that each Raga is ideally performed during a specific period of the day or night. The “time-theory” is controversial. But, the micro-level classification has a psycho-acoustic logic that justifies independent attention.


Treating it as a psycho-acoustic classification that respects the tonal character of Hindustani music, I have enlarged the number of categories to be a more comprehensive coverage of the psycho-acoustic typologies encountered.


It must be pointed out that these are not necessarily mutually exclusive or even collectively exhaustive categories. Some Raga-s may belong to more than one category, and some others may belong to none of the categories and may be classified as “Others” in Table 3C.


(a)  Raga-s deploying Komal (flat) Re (2nd) and Komal (flat) Dh (6th): The two swara-s are in first-fifth correspondence, with a pitch ratio of 1:1.5.


(b)  Raga-s deploying Shuddha (natural) Re (2nd) and Shuddha Dh (6th):  The two swara-s are in first-fifth correspondence, with a pitch ratio of 1:1.5.


(c)   Raga-s deploying Komal (flat) Ga (3rd) and Komal (flat) Ni (7th): The two swara-s are in first-fifth correspondence, with a pitch ratio of 1:1.5.


(d)  Raga-s deploying Shuddha (natural) Ga (3rd) and Shuddha (natural) Ni: The two swara-s are in first-fifth correspondence, with a pitch ratio of 1:1.5.


(e)  Raga-s deploying Tivra (sharp) Ma (4th): This category has independent psycho-acoustic significance because, it does not find acoustic correspondence (1/4 or 1/5) with any other swara of the standard 12-swara scale of Hindustani music.


(f)    Raga-s deploying both, the Shuddha (natural) Ma (4th) and Tivra (sharp) Ma (4th). This is a rare category in Hindustani music. Because of the “pivotal” character of Ma (4th) in the scale, this category is of special psycho-acoustic significance.


From the analysis of the top-42 Raga-s, it emerges that Raga-s deploying Komal Ga and Ni and those deploying Shuddha Re and Dh dominate the composite index. These two categories happen to occur most typically in Raga-s performed after sunset. This is an expected pattern considering the most frequent timing of stage concerts. This, however, also emphasizes an unintended – and surprising -- reality.


Despite the substantial and growing role of remote media (recordings and online archives) as carriers of Hindustani music, the “Time Theory” bias relevant to “personal” (direct encounter) exposure still appears to significantly shape the Ragascape. This is probably because YouTube content is still, primarily, a byproduct of stage performances and festival events. A diversification may begin to take place once the Hindustani music content starts getting created especially for online transmission under assumptions of time-neutral access by audiences.


This process has gained some traction during the period when Covid-19 suspended concert activity for a long period. Besides YouTube, Facebook also emerged as a significant vehicle for online delivery of Hindustani music to audiences in video format. As the world crawls towards “normalcy”, there is an expectation that, because of its essentially interactive nature, Hindustani music will soon return to the concert platform as the preferred and more influential – though not exclusive -- interface between the musician and the audience.

 However, there is also a contrarian expectation – that the attractive economics and superior convenience of the online alternative will support the movement towards in-house/ studio production and online delivery. Such a drift could also be encouraged by the growing attraction of the international market for concerts as well as online teaching opportunities for Hindustani musicians. The future is impossible to foresee, and its implications for the Hindustani music ecosystem are difficult to envisage.




Based on the results of the three preceding studies, and their integration into a composite view of the Ragascape phenomenon, we do have here an answer to the issue raised by Bhatkhande’s classification postulated a century ago. It appears that there is, indeed, a group of 40/50/60 Raga-s, which constitute the active “core” of the musical culture. Comparing our contemporary list of the dominant Raga-s with Bhatkhande’s century-old list is neither attempted, nor relevant.


Our inference tangentially confirms the observation that most leading musicians of each generation have built their individual careers focusing their energies on a finite set of Raga-s (may be, 15 to 20), which collectively build this core of 40/50/60 Raga-s. But, this numerical inference probably under-estimates the usefulness of this series of studies.


Departing from Bhatkhande’s binary classification of Raga-s into “Prasiddha” and “Aprasiddha”, our study has adopted a calibrated view of the Raga spectrum, which is closer to the perspectives of market research. Each Raga considered here is measured on the dual criteria of availability and acceptance. In a sense, each Raga represents a “market” -- because we have accounted for both, its availability and its acceptance. What we have, as a result, is a “Market Rating Index” similar to the ones consumer product marketers use for determining their marketing strategies. Different marketers use the same Market Rating Index and come up with a marketing mix and market focus uniquely suited for their own distinctive competencies and marketing goals.


This study has attempted to build a Rating Index for 81 Raga-s out of the 97 candidate Raga-s selected originally on the basis of a reasonable criterion of inclusion. The results obtained are, admittedly, a product of the data-source I have chosen to rely upon, and the analytical approach I have chosen to adopt. Any alternative data-source, analyzed differently, can deliver totally different results.  The relative validity of any number of alternative methodologies will, of course, remain impossible to assess because of the amorphous nature of this phenomenon.


Every data-source, or even a multiplicity of data sources, can be questioned for  “representativeness”. Every analytical approach can be questioned for its suitability. None of these limitations will, it is hoped, deny the usefulness of this endeavor to the music community.


© Deepak S. Raja. February 2021.


Monday, October 19, 2020

The Ragascape of Hindustani Music: VIII


Contemporary Musicians and the Ragascape


In this third part of the Ragascape study, we examine the repertoire strength of different Raga-s, as evident from the concert repertoire of contemporary musicians. This direction of enquiry is suggested by the findings of the first study on Audience Engagement Indicator of widely performed Raga-s. (See Parts 1 through 6)  In the earlier study, the recordings of contemporary musicians reported a much higher Audience Engagement Indicator than the recordings of earlier generations of musicians. This justifies a generation-specific enquiry related to the repertoire-strength of Raga-s.


The relationship between the Ragascape as a cultural umbrella of melodic ideas and the repertoire of contemporary musicians will tend to be reflected in two ways. (1) It will be reflected in the totality of Raga-s performed by prominent contemporary musicians and (2) It will be reflected in the frequency with which each of the Raga-s is performed. It is impossible to access information on the frequency with which each Raga has been performed in public by prominent contemporary musicians. However, it is possible to get some indication of the totality of the Raga-repertoire being performed by contemporary musicians. I choose to draw, once again, on the YouTube archive to paint an indicative picture of this phenomenon.


Defining the Contemporary Musician


For the purposes of this study, I define the contemporary musician as a vocalist or instrumentalist who is born between 1950 and 1980, and is still active as a professional musician. The logic for this definition is three-fold. (1) The birth in/after 1950 defines the first generation of post-independence musicians. (2) Social scientists accept a period of 30 years as appropriately defining a cultural generation. (3) At the time of doing this study, a musician born before 1980 is at least 40 years old which is considered the maturation threshold of musicians in the Hindustani tradition.


Based on this criterion, I drew up a list of 20 prominent vocalists and 20 prominent instrumentalists. Every single Raga performed by them anywhere, on any occasion, and uploaded on YouTube was logged. Only solo performances were considered, because the choice of repertoire in duets or ensembles need not reflect the musician’s own preferences. All major genres of Hindustani music were covered by the survey – Khayal, Thumree, Dhrupad.


The output table answered the question: How many musicians have performed a particular Raga even once, irrespective of context and location? As mentioned earlier, no estimate is possible of how frequently each musician has performed a particular Raga, and therefore this factor remains unquantified. The list of contemporary musicians covered is given in Table 1, along with their years of birth, and instruments (where relevant).

  Analysis of results


An interesting figure emerging from this study is the difference between the number of Raga-s performed by vocalists and instrumentalists. Vocalists appear to present an average repertoire of 44 Raga-s, while, instrumentalists appear to perform an average of only 35 Raga-s. No significance can be attached to this difference, because it may well be an infirmity of the data-source without having any basis in the reality. In any event, the “finding” is irrelevant to the objectives of this study.


Tracking the YouTube upload repertoire of the selected 40 musicians gave us a total of 244 Raga-s. Of these, 110 have shown up as having been performed by both vocalists and instrumentalists, 43 performed only by instrumentalists, and 91 performed only by vocalists. Because the total Repertoire-strength of different Raga-s is our focus, I choose to analyze only those 110 Raga-s which have been performed by both, vocalists and instrumentalists.



The frequency distribution of Raga-s against the number of musicians, and the limitations of the data-source, will not permit sophisticated statistical inference. However, since we are interested in the dominance of Raga-s performed by the largest number of musicians, we may focus our attention at the upper end --  Raga-s enjoying a larger presence on account of the number of musicians performing them.


An analysis of 110 Raga-s performed by instrumentalists as well as vocalists reveals an average of 12.4 musicians (out of 40) having performed any Raga, with a Standard Deviation of 7.9. This simple measure permits us to focus our attention primarily on those Raga-s which have been performed by more than 12 (30%) musicians.


This group of Raga-s enjoying an “above average” preference amongst musicians can be further divided into two groups. Group 1 is of those Raga-s which are performed by more than 20 out of 40 musicians (above mean+ standard deviation), and Group 2 performed by 12 to 19 (Mean to Mean + Standard deviation) musicians.


Group 1 consists 0f 20 Raga-s, and Group 2 consists of 33 Raga-s. The Repertoire-strength of these 53 Raga-s is given in Table 2 at the end of this paper. These Raga-s may be said to constitute the Ragascape focus, by virtue of being performed by the largest number of prominent contemporary musicians. A contextual/ thematic classification of the two groups is as follows.

The two groups are to be seen as a continuous ranking of Raga-s on the basis of the number of musicians performing them. They do not represent either qualitatively different groups, nor do they represent different sets of musicians. They simply represent "Very high popularity" and "Moderately high popularity" ranges. 


A comparison between the high-rank Group 1, and the intermediate rank Group 2 is interesting. Group 1 has a proportionately larger presence of late evening Raga-s than Group 2.Carnatic Raga-s fall entirely in the intermediate popularity region.  Late-evening Raga-s falling more predominantly in the high-popularity region is easy to explain. Many more public events are held in the late evening than any other time of the day.  Carnatic raga-s  fall in the intermediate popularity region perhaps because fewer musicians have studied them well enough to perform. 


Having ranked Raga-s on the basis of their presence in the repertoire of contemporary musicians, it is useful to examine the correspondence between this ranking and the Audience Engagement Indicators derived from YouTube viewership in the first part of this three-part study.


Contemporary Repertoire and Audience Engagement Indicators


This present study has identified 110 Raga-s for focused consideration, and ranked them on the basis of the number of contemporary musicians performing them. The earlier study started from the other end – considered 97 commonly performed Raga-s, and ranked them on the basis of an Audience Engagement Indicator.  A matching of the two lists shows 89 Raga-s appearing in both the lists. A statistical correlation was run on the 89 common Raga-s to examine the correspondence between the two rankings. The  procedure yielded a Correlation Coefficient of 0.65, and the X-Y scatter plot is presented below:



The plot exhibits a substantial correspondence between the dominant repertoire of contemporary musicians, and Audience Engagement Indicators. However, it also suggests that the collective repertoire of prominent musicians is not entirely guided by Raga-s generally believed to be popular. To this extent, it may be said that contemporary musicians are manifesting the Ragascape of the present, while also shaping the Ragascape of the future by exposing contemporary audiences to Raga-s that may not belong to the known category of popular/ populist Raga-s. 

This observation is, of course, defensible only as an assessment of the collective tendencies of contemporary musicians. Within this collective phenomenon, there are musicians who perform only the popular/ populist ragas, those who also perform Raga-s of intermediate popularity, and those who perform Raga-s they want/like to perform irrespective of their popularity.


Concluding observations


In these three studies, examining the concept of a Ragascape from different angles, we have established that there does exist a set of Raga-s that can be considered the core of the present-day musical culture, and that its existence is manifested in several aspects of musical activity. Admittedly, the data-source utilized by us to arrive at these inferences does not satisfy the demands of scientific rigor. However, the ecosystem we seek to study is, by its very nature, inhospitable to conclusive research. This series of studies has attempted to use YouTube, the only candidate data-source,   as prudently as possible to evolve a plausible snapshot of the cultural phenomenon sought to be profiled.


These three studies provide us with the insights that, if merged, may lead to a Raga-repertoire profile of Hindustani music that may answer to the definition of a Ragascape. The results may, or may not, approximate Bhatkhande’s definition of “Prasiddha” Raga-s, as conceived by him over a century ago. It could, however, answer a similar query, with a more transparent support of evidence than was possible in his time.

(c) Deepak S. Raja. October 2020

Thursday, September 10, 2020

The Ragascape of Hindustani Music - VII

The idea of the Ragascape has been discussed in the earlier part of this study.  The testing continues now, though from a different angle, of the hypothesis that the contemporary Ragascape is dominated by a finite number of Raga-s, and that the number could be in the region of 50.

The Ragascape manifests itself in a wide variety of personal and impersonal encounters between musicians and their audiences. Amongst personal interfaces, annual music festivals held in different parts of the country – and now also in London – are significant.

Music festivals are, typically, events lasting anywhere between 3 and 13 consecutive days, in which the organizers attempt to feature musicians of national stature, along with distinguished local musicians of the region, and sometimes also promising local talent. As such, they attract the attendance of the entire local music community, ranging from professional musicians and connoisseurs, to lay music lovers. The major festivals also tend to be reviewed in the local media. Most music festivals in India are held during the winter and spring (December to February), when the weather is clement, and open-air seating arrangements can be made to accommodate large audiences, which are known to range from a few thousand to 10,000. These factors make annual music festivals a potent source of insight into the musical culture.

Does this source lend itself to intelligent observation and interpretation? How does a researcher observe and interpret the content of music that has been performed over several decades at different locations – not only in India, but now also abroad? An approach made for historical data to two of the better organized festival hosts found them either unable or unwilling to help. This phenomenon is endemic to the ecosystem that manages Hindustani music events in India, and is rooted in the history of the major festivals, one of which has been organized annually for over a 140 years.

The background

The organization of annual music festivals started in the late 19th/ early 20th centuries when feudal patronage of classical music was fading. By this time, the radio and commercial records had begun to cultivate a large “public” audience for classical music. However, neither radio sets nor commercial records were cheap, and the direct/live experience of classical music was still beyond the reach of the average music lover. At that stage, mercantile patrons stepped in with funds, domain knowledge and organizational competence to bring face-to-face musical encounters to the general public through annual festivals.

The growing popularity of music festivals in the host cities resulted in a growing need for funds and organizational competence. Over a period, individual philanthropists either promoted, or were replaced by, non-profit/ charitable trusts, with funding solicited from a multiplicity of donors. This transition did not, however, professionalize event management activity for three main reasons: (a) The basic orientation remained philanthropic/ non-profit with audiences either not paying at all, or the box office contributing only in small part to the total cost of the event. (b) The primary objective of the events has remained to service the local music community of the city with a live musical experience. (c ) A purely seasonal activity could not justify a full-time establishment to pursue emerging artistic, technological, and economic trends and opportunities.  

The recording of festival music arrived perhaps in the 1960s, when quality tape-recorders became accessible on a wide scale at a reasonable cost. It was about then that organizers began to make audio recording of festival music. Much of this asset remained the private property of the organizers and circulated invisibly in the grey market. Some organizers established tie-ups with music companies to duplicate festival recordings, and market them on a commercial scale. Thus began the era of festival music (a) being delivered through impersonal storage media as a product and (b) reaching audiences living beyond the confines of the host cities. The financial implications of this commercialization for the event organizers are opaque, while also being irrelevant because very few festival hosts could contract such arrangements. 


The filming of festival performances was very rare until the video-tape era arrived. In the early stages of videography, filming was discouraged by costs of production as well play-back and an unexplored   market for copies. The costs of videography and related storage-media became interesting only with the arrival of the video CD at the turn of the century. But, hardly any major festival organizer began marketing video CDs of festival music. Upon the arrival of the Internet, and specifically of YouTube, the incentive for the video-preservation of festival music took on a more interesting turn.  

YouTube permits the monetization of uploaded assets. In addition, a higher level of visibility for the event organizers through “Publish Yourself” media could stimulate the sourcing of sponsorship funds to support the festival activity. For festivals which rely even partially on box office collections, a higher visibility in the social media offered the possibility of attracting larger audiences for future events. With the arrival of amateur videography equipment, musicians featured at the music festivals found it possible to upload their own performances on the social media to enhance their visibility in the concert market.

These attractions, working collectively after the dawn of the 2nd millennium, have brought a substantial part of festival music online through YouTube. However, the basic culture of most major festival organizers remains stuck in the conventional mode, focused on live music delivered to a local music community. The information needs of a global audience accessing music through a remote medium do not appear to be of concern to them. As a result, a comprehensive, and professionally catalogued archive of festival performances is difficult to find on YouTube.

With all its imperfections, the archive presently available on YouTube holds some promise as an indicator of the Ragascape of music festivals, providing a different perspective on our search for an understanding of the musical culture. In this belief, I surveyed YouTube uploads of the following music festivals, and have attempted to extract some meaning from the results of the survey.  

The survey is not a census. It is not based on a sample of any kind – whether systematic or unsystematic. It utilizes whatever was available on YouTube, and could be identified as eligible for inclusion.

The coverage of festivals

The Harballabh Sangeet Sammelan covered in this survey is the oldest, having been held consecutively for 146 years, while the Choudhry House Festival held in Kolkata is the youngest, having commenced as an organized public event in 2012. Wherever the event organizer is more than 20 years old, and the year of the performance is available, I have chosen to log only performances of Raga-s performed since the year 2000.

No particular criterion has been applied to the selection of festivals for this survey. The selection was guided by the availability of a sizeable inventory of eligible observations on YouTube. An attempt was made to select significant festivals reflecting the musical cultures of different parts of the country. The Darbar Festival in London was added because it represents a significant presence of Hindustani music in the largest foreign home of the Indian/ South Asian community.

      The method of survey is simple. Identify YouTube videos displayed as being of a performance recorded at any of the surveyed music festivals, ensure that the performance is less than 20 years old in August 2020, and list the Raga performed. No distinction is made between vocal and instrumental performances, nor on the basis of the age of the musician performing a Raga. These refinements would be interesting. But, the quality of the data-source may not have supported them. Therefore, the only feature of relevance considered is the Raga performed.

The operative measure of a particular Raga’s significance -- in this study --  is the number of times it occurs in all festivals collectively across the surveyed recordings. Considering the limitations of the data source, no greater refinement seemed inviting. I regard the data as more suited for qualitative, than quantitative, interpretation. Some numeracy, however, helps highlight the phenomenon we seek to understand.  

Results of the survey

The survey covers 7 music festivals. The results are based on an audit of 591 performances, spread over a period of 20 years, in which a total of 132 Raga-s are encountered. The analysis of the frequency with which each raga occurs in the survey reveals a median of 3 occurrences, a modal value of 1, and a mean of 4.8 occurrences with a standard deviation of 4.5. The distribution of occurrences is highly skewed (see graph). 

This pattern obliges us to keep the argument simple. To obtain a meaningful isolation of the extremeties, I have divided the 132 Raga-s into three groups.

 Group 1: Raga-s occurring five times or more in the survey i.e mean or higher, 

Group 2: Raga-s occurring above the mode, but below the mean i.e. 2 to 4 times

Group 3: Raga-s occurring at the mode, i.e. only once. See Tables of Group 1,2, and 3 at the end of this report.

This approach divides the Raga-s into three approximately equal groups of +/- 40 each.

Group 1 Raga-s

Group 1 (Raga-s occurring five times or more) consists of 44 melodic entities, which appear to be the “standard contemporary repertoire” of Hindustani music in the context of music festivals. The most frequently performed Raga in this group (Bihag) occurs as many as 23 times in the survey.

It is useful to consider how this group of 44 Raga-s dominating the festival circuit compares with the Top-44 Raga-s identified by us through the measurement of an Audience Engagement Indicator from YouTube viewership data in the earlier study. Only 28 Raga-s were found to be present in both the Top-44 groups. The analysis of the common 28 Raga-s yields a Correlation Co-efficient of 0.4. (see Graph below). The Table showing the comparison is given at the end of this report.

There are two facets to the interpretation of this correlation. Firstly, the dominant repertoire tends to partially anticipate the comfort levels of audiences. Secondly, the specific context of a music festival encourages certain choices which cause festival repertoire to diverge from the assumed “comfort zone” of audiences. Both these facets are reflected in this analysis, despite the far-from-ideal quality of the data source.  

In interpreting this analysis, we need to remember that the YouTube Audience Rating reflects a remote global audience, while the festival repertoire of musicians addresses a physically present local audience. To this extent, the two data series may not, strictly, be comparable. The geographical component of the comparison cannot be isolated. The comparison does, however, highlight the specific context of a music festival in some respects.

 A contextual/ thematic classification of the two Top-44 lists is interesting. (see table below).

The classification of the Top-44 Raga-s from the two studies shows a broadly similar pattern, except in three cases. Early morning Raga-s appear to engage YouTube audiences more than musicians appear to offer them at music festivals. The most likely explanation for this divergence is the time-specific prescription related to the performance of Raga-s. All festivals do not host morning sessions – though some do -- and therefore the opportunities for performing appropriate Raga-s for that time of the day at festivals are scarcer. The higher occurrence of sunset Raga-s at the festivals also reflects the same reality. 

A similar logic may explain the plentitude of night Raga-s at music festivals compared to their popularity among YouTube audiences. Music festivals in India often stretch late into the night, and sometimes into the early morning, providing a wider canvas for night raga-s to be performed. Another facet of this phenomenon is that “star musicians” at music festivals are generally featured at the end of the evening session, with no time-limit set on their performance. This convention can automatically enhance the presence of night Raga-s.

 While the patterns at the top end of the scale is largely predictable, the pattern at the bottom end is interesting for its enigmatic quality.

Group 3 Raga-s (Single occurrence)

As many as 15 of the 42 Raga-s occurring only once in our festivals survey are undocumented. For determining their status, I rely on Raga Nidhi, by B Subbarao (Music Academy, Madras), which has documented the largest number of Hindustani and Carnatic Raga-s. Remaining undocumented implies that they are either extremely rare Raga-s, or modern/ contemporary creations awaiting documentation. The occurrence of such Raga-s at the bottom of the scale reflects the scope musicians have, in the music festival context, to pursue novelty, while risking audience discomfort. 

It is interesting that the bottom end of the scale does not feature a single “Thumree Raga”. This category is over-represented at the upper end of the scale.  Interestingly also, the bottom end (42) shows a higher presence of seasonal Raga-s (5) than the Top-44. On closer scrutiny, this group is found to include compound Raga-s, like Paraj-Basant, and Miya-ki-Sarang, which are rare anyway. Of course, the group also includes the widely performed and heard Gaud Malhar and Basant-Bahar, whose presence at the bottom end cannot be explained except as an infirmity of the data-source.  

 The YouTube Rating 97 and the Festival 139

 Based on the top-44 Raga-s emerging from the two studies, we are allowed to believe that the partial convergence, and substantial divergence between the two lists of dominant Raga-s. The divergence also appears to have some logical explanations related to the context and methodology of the two studies.

A different view emerges if we consider the two lists in their entirety. We have 97 Raga-s in the Youtube Audience Rating study, and 139 Raga-s emerging from the Festivals study. It is found that the two lists have 75 Raga-s in common. A graphic plot of the correlation between the 75 common Raga-s is shown below:


The Correlation Coefficient of 0.61 is considerably stronger than the Top-44 coefficient. Here, we can witness a more comprehensive reflection of the cultural process. This further supports the hypothesis that the Raga-s populating  the Ragascape of the era, creates the semblance of an “unstated agenda” between musicians and their audiences. The precise number of Raga-s in this unwritten contract is not important. What is important is that the number could be finite, and a small fraction of the total of 1200+ documented Raga-s.


The earlier study of Audience Engagement Indicators on YouTube gave us a list of 30 Raga-s which could be considered the “core” of the Ragascape we are attempting to map. This study, based on an analysis of music festival repertoire, gives us 44 Raga-s which appear to constitute the “standard festival repertoire” on the contemporary Hindustani music scene. There is, expectedly, considerable duplication between the two lists. However, this study has also permitted us a broader view of the cultural process that brings the identified Raga-s into dominance. Having considered these two approaches to the mapping, we have come a little closer to achieving the objective of this study. Further approaches can be considered for a more categorical mapping of the Ragascape. 

(c) Deepak S. Raja, August 2020

Tuesday, June 16, 2020

The Ragascape of Hindustani Music: VI

Methodology and the data source

Research design for this study had first to determine the set of Raga-s to be considered.  The list of “Candidate” Raga-s was compiled by merging the under-graduate and post-graduate syllabi of the major teaching universities and examining bodies granting certifications/ degrees/ diplomas in Hindustani music. This list was submitted for suggestions/ approval to a panel of musicians known for their vast Raga repertoire and academic orientation. Through this two-tier process, a list of 97 Raga-s was finalized, for which audience engagement measures were to be computed.

For compiling the data, the YouTube platform was searched for each Raga. The coverage of “significant” musicians performing each raga was attempted as a census (100% coverage), and not as a sample.  The assessment of “significance” utilized my own knowledge of musicianship over the last century on the national as well as regional levels. Considering the global scope of the enquiry and the diversity of material, my judgement of “significance” could have been less than 100% defensible. Care was taken to include significant musicians representing (i) vocal music and all significant instruments (ii) Dhrupad, Khayal, and Thumree genres (iii) contemporary/ fusion genres where recordings explicitly claimed and performed Raga-based music (iv) musicianship originating in Pakistan, Bangladesh, and Afghanistan performing Raga-based music (v) musicianship originating in the US, Europe with credible performing attainments (vi) musicianship originating from the Carnatic tradition but performing Hindustani/ shared Raga-s in Hindustani style.

Media researchers have found that YouTube on-screen data is singularly unhelpful for systematic research. This would seem intentional because unambiguous data can be used for competitive purposes (can help some musicians at the cost of others). This can jeopardize YouTube’s impartial status as a media owner/ platform provider, and damage its commercial interests. Having conducted a major study in 2015 based on YouTube data, (Refer Chapters 15 through 20, in The Musician and His Art – Essays on Hindustani music: DK Printworld, New Delhi 2019) I had developed a method of usable data acquisition from YouTube. This method obliged me to personally create every one of the 8100 records for analysis. Despite utmost diligence in database creation, I am unable to vouch 100% for the soundness of my results. YouTube’s evasive purpose is fulfilled.  

A notional limit of 100 data-points was envisaged for each Raga. The data-base compilation was completed in the last week of March 2020 and the first week of April 2020.

The Audience Engagement Indicator has been computed as follows:

 Limitations of YouTube data

YouTube as data source is far from ideal for systematic research, and that I am using it in full awareness of its limitations. Despite this, it is now being used by researchers even in the more advanced research environments. Presumably because it has usable value, its use will not stop; and because of its known limitations, the controversy over its value will not end.

The perception of limitations will obviously be guided by the researcher’s objectives – what he is looking for, and what he may instead be measuring. My concerns are focused on the following infirmities.

YouTube reports “views” for each upload. The basic question is: what is a view? What minimum duration of exposure qualifies a “visit” as a “view”? The information displayed on the screen does not distinguish between a 10% viewing of a video and a 40% viewing of it. The difference matters to me.

The duration aspect of viewership is also connected to a change in YouTube policy some years ago. There was once a duration limit on uploads. This affected Hindustani music very significantly, as most performances exceeded the limit, and had to be split into 2, 3, and sometimes 4 parts for upload. After the duration limit was lifted, mostly complete performances were uploaded. Looking at that data today, we implicitly equate the engagement measure of a concert split into 3 parts with that of a complete performance. This seems unreasonable.  

YouTube viewership data is cumulative from the date of uploading. By using that data, I am implicitly accepting one view of 2010 on par with one view of 2018.  Intuitively, we know the assumption is flawed. During this period, YouTube content has changed substantially, viewership has grown exponentially, and audience profile for every kind of content has almost certainly changed radically. No correction factor can help.

Another problem with cumulative data is that it effectively equates 20,000 views accumulated over 20 months with 40,000 views accumulated over 40 months. Intuitively, this equation does not look reasonable. If the propensity of a recording to accumulate viewers is important, 40,000 over 4 years is more valuable. And, if the speed of audience accumulation is considered important, 20,000 over 20 months is more valuable. YouTube data is unhelpful in this respect.

The YouTube audience is global, and so is the audience for Hindustani music. But, we have no data on the geographical spread. By implication, we are accepting that foreign audiences of Hindustani music and foreign musicians —across all nationalities and cultures -- have the same relationship with the music, as Indians have. This is unrealistic. The opacity of the geographical spread can easily mislead us – as it has the possibility of doing in this study.

YouTube viewership – by whatever criterion registered – is a partially manipulated expression of audience engagement. This is because a sophisticated program guides the viewer into “viewing” content beyond his purpose, and spending much more time on YouTube than was necessary or planned. As a result, the numbers we see include “incidental” viewership. A separate reporting of the primary (search word) and incidental viewership would be very helpful.   

Although YouTube is a video medium, the nature of the content is itself not uniform. In fact, even the notion of “viewership” may be irrelevant to a lot of the content. Do 100 people listening to an audio recording with just a photograph of the musician on the screen represent the same level of audience engagement as 100 people watching him or another musician in action on film? If not, how much can we depend on a standard measure of audience engagement across different content formats?

With specific reference to Hindustani music, YouTube neither offers a uniform media experience to its audience, nor publicly provides a rigorous measurement of audience engagement. What, then, does YouTube data provide? It cannot be said to provided “statistics”, but can suggest “orders of magnitude”. Inferences can be drawn judiciously from its analysis, with every inference reflecting the analyst’s awareness of data limitations.

YouTube has invested heavily in generating analytics for user management and advertising value assessment. At some stage, it will have to start understanding itself as a cultural force. It will then benefit by cultivating communities of media researchers through a more transparent stance with respect to its data assets.

Until this happens, the Indian musicologist should be content with indicative inferences. Is this better than the “Delphi method” of polling 10 veteran connoisseurs and observers of the music scene? I believe so because, firstly, surveying this population globally is an almost impossible task, and secondly, present-day oracles – whether of Delphi or anywhere else -- are susceptible to "personal" preferences, biases and prejudices; impersonally generated numbers are not.