False Memories and Dissonant Truths: Digital Newspaper Archives as a Catalyst for a New Approach to Music Reception Studies

by Christopher Dingle and Laura Hamer

1. Introduction: A Golden Age?

One morning, before too long, you are going to wake up and find last night’s opera premiere reviewed in your newspaper by Covent Garden’s Chief Executive and the new play at the National by a drizzle of audience comments. The role of the professional Arts critic is being eroded. […] American newspapers are shedding critics as the first line of economy. In Britain, review space has shrunk […]¹

Recent years have seen much conjecture regarding the state of classical-music criticism, and how it has changed over the second half of the twentieth century. According to, amongst others, Norman Lebrecht, the specialist music critic is on the verge of extinction, newspaper editors no longer value classical music, and coverage of music has diminished in both quantity and quality. Discussion in classical music circles often centres upon claims of a profound change in priorities from a generation ago, with a general conviction that space, which once would have been dedicated to classical music, has been usurped by popular genres. Edward Rothstein, in his New Grove article on ‘Criticism: Since 1945’, articulates wide-spread beliefs about changes in the nature of criticism since the Second World War, and the perceived threat of the rise of popular-music criticism when he claims that: ‘criticism changed dramatically during this period […] By the end of the century, the main critical controversies no longer concerned new music or performance styles, but whether classical music had a greater claim on any culture’s attention than any other forms of music and entertainment.’² Rothstein reinforces his comments regarding classical music’s loss of an a priori privileged position by reference to the development of a pluralistic musical culture: ‘readers could no longer be assumed to share a similar background of musical experience. Pop music and world music became more dominant, challenging the boundaries and claims of classical-music criticism’.³

However, to what extent can we take declarations such as Lebrecht’s and Rothstein’s at face value? Is music criticism in decline? Is the specialist music critic an endangered species? Does classical-music criticism really have to share its coverage more and more with other forms of musical expression? Recent discussion of classical-music criticism all too often centres upon anecdotal claims, but comparisons supported by verifiable evidence have thus far been difficult to make. Digital newspaper archives now raise the possibility of new ways to interrogate the actual evolutions and revolutions of music criticism since the Second World War by replacing speculation with facts supported with rigorous evidence. The British Music Criticism since 1945 Project (henceforth BritCrit Project) aims to exploit these archives to the full by carrying out such a thorough interrogation of the actual changes which have taken place within music criticism in British newspapers since 1945.

One of the strongest reactions which the authors have encountered when discussing this project with a wide range of colleagues, both within and beyond musicology, has been a wide-spread belief that standards and outlets for music criticism are in decline, that the role of the music critic is under threat, and (perhaps most fervently) that an increase in popular-music criticism has had a detrimental effect on that of classical music by eating up the latter’s space. Whilst such responses can only ever be classed as anecdotal evidence, they do appear to reveal deep-seated cultural impressions of decline, and of the threat of popular-music criticism. Gloomy reactions to today’s criticism, furthermore, are often accompanied by nostalgic evocations of a (ill-defined) golden age of music criticism – located a generation or two before the present – when daily broadsheets allegedly carried lengthy essays which were dedicated exclusively to classical-music criticism. Systematic investigation of classical-music criticism in British newspapers since World War Two has, thus far, however, failed to uncover such a zenith. Despite factors such as the undoubted growth in popular-music criticism and more recent democratisation of comment, our initial research indicates that classical-music criticism coverage has not experienced a precipitous decline in British newspapers since the Second World War and challenges the notion of a ‘golden age’ of music criticism.

The BritCrit Project hosted by Birmingham Conservatoire, through developing a new methodology for investigating the vast repository of music criticism preserved in digital newspaper archives, aims to replace speculation by fact. Blending traditional musicology with techniques drawn from information technology, corpus and computational linguistics, journalism studies, and historical geography has enabled an approach whereby qualitative research on the changing trends in British music criticism is supported by quantitative, statistical results. Our primary research question addresses: how has coverage of music in newspapers changed since World War Two? This can be subdivided into two supplementary questions, which themselves have prompted a raft of further queries:

• What importance is attached to coverage of music in newspapers since World War Two?
• What qualitative changes (if any) have there been in the coverage of music?

The first of these supplementary questions is closely related, but not directly analogous, to ascertaining how much space is devoted to music. As will become clear, what appears to be a simple quantitative question is in reality fraught with methodological difficulty. The focus of this article is the initial stages of the project, which concentrated upon The Times newspaper from 1945 to 1985 with some sampling from later issues and also The Guardian/Manchester Guardian. The article discusses some of the challenges involved in charting content across large timescales before outlining the methodology devised for this project, and the advantages that it presents for musicologists.

2. In Search of a Methodology

It might be thought that charting the importance attached to music within newspapers would be a relatively simple task. Surely, this ought to be a simple quantitative matter of measuring the coverage devoted to music in, say, 1962 and comparing that with the figure for 2012. In fact, making a comparison of the importance attached to music (or any other subject) in newspapers either across eras and/or between different titles is a complex undertaking. There appears to be no established coherent methodology for quantitative mapping of material in newspapers across such broad time frames. The reasons for this initially surprising omission soon become clear, with caveats at almost every turn. The problem of data gathering provides the first hurdle.

‘Column-inches’ is the term commonly associated with newspaper content in popular culture. For the purposes of examining the make-up of newspapers, this beguiling, apparently quantitative term is actually meaningless beyond individual titles within specific timeframes. It is a handy signifier of the constant need for content in newspapers, in the same way that a Bunsen burner is the standard visual symbol for scientific research, even though very few scientists actually use Bunsen burners in their work. The definition of a column-inch is simple, being a space that measures one inch by the width of one newspaper (or magazine) column. The earliest citation in the Oxford English Dictionary dates from 1940,⁴ replacing the earlier ‘Square’ in the US, and the term is most commonly used in relation to advertising. The column-inch is an effective content-measurement tool for a single issue of a single newspaper, or (usually) contemporaneous issues from the same newspaper. However, it is unreliable for comparative content analysis, not least as there is no standardised measurement for column width. The number of columns on a page varies from newspaper to newspaper, and across time periods. In 1951, The Times had seven columns and the Manchester Guardian eight. Both The Times and the Guardian split their pages into seven columns in 1961, but eight columns in 1981 and in 2011 five columns. Each change was accompanied by a change in column width, but this is just one fluctuation. Measuring content in newspapers is also undermined by the regular changes in font, font-size, layout and size of page. The parameters of the text in an issue of The Times from 1961 will be different from those in 2011, and may be different from those in 1966 or 1971. They may also diverge from the parameters used in the Guardian and The Telegraph. In other words, a column-inch of text in a newspaper from 1961 may contain a different amount of content to one from the same newspaper in 2011, or a column-inch of its competitors at any time. In attempting to assess broad trends across decades, these variations can rapidly become statistically significant.

It is now common practice, thanks to word processors, to measure length in terms of word or character count. Getting this figure for any given article is not especially problematic. For current newspapers an electronic edition can be accessed, either through online versions or via the Nexis UK database. For older newspapers, the advent of digital archives has helped, though not all provide word counts for articles. Even before such resources, it was possible manually to provide an actual or estimated word count. However, the figures produced are of benefit only for direct comparison of article length. On their own, they do not give any sense of how this relates to the content of the newspaper as a whole. For the word count of individual articles to have any contextual meaning, the word count of the newspaper as a whole needs to be known, and possibly figures for the subsections within it. At this point, all digital resources fall short. Online editions of newspapers are not identical to the print versions. Even estimating a figure for the latter is cumbersome, while the notion of a daily edition is questionable for online versions and, even if it was not, a total word count could only be made by collating information from each of the individual articles. Digital archives of older newspapers have been created on the basis of the article. This understandable method for scanning and organising the content again means that the task is extremely time-consuming.

Another possibility is simply to count the number of articles, possibly refining this to a very rough categorisation in terms of size. For instance, a comparison can be made for music content on equivalent Fridays in The Times as in Table 1:

Table 1: Samples of classical music content on Fridays in The Times.

Date*		Content†
		Articles	Reviews Concert/Recording		News Pieces	Miscellaneous
1950s	6 April 1951	1	2		¼
	13 July 1951		2
	5 September 1952	1	1		1
	24 April 1953	1	1		¼
	11 December 1959	1	1		3
1980s	10 April 1981		3		½	1 book review
	17 July 1981	1	2		1
	10 September 1982		3		1
	29 April 1983		2
	14 December 1984		3
2000s	11 December 2009	1	3
	8 April 2011	1	1	1		8 entries (out of 29) in Festival planner
	15 July 2011	2½	1	1	1
	7 September 2012	1½	3	1		1 TV preview
	26 April 2013		1	1

* all Fridays

† fractions indicate brief pieces

On this basis, this very small random sample suggests, if anything, an increase in content from the 1950s to current times. Moreover, these figures only count classical content, and the numerical increase is even starker when all music is included, for jazz, popular, folk and world music reviews articles and news stories did not appear with any regularity at all in the 1950s, and were regular, but limited in the 1980s, as can be seen in Table 2:

Table 2: Samples of all music content on Fridays in The Times.

Date*		Content†
		Articles	Reviews Concert/Recording		News Pieces	Miscellaneous
1950s	6 April 1951	1	2		¼
	13 July 1951		2
	5 September 1952	1	1		1
	24 April 1953	1	1		¼
	11 December 1959	1	1		3
1980s	10 April 1981		3		½	1 book review
	17 July 1981	1	3		1
	10 September 1982		3		1
	29 April 1983		3
	14 December 1984		3
2010s	8 April 2011	8	3	1		21 entries (out of 29) in Festival planner
	15 July 2011	3½	5	3	1	1 obituary
	7 September 2012	3½	5	2		2 obituaries; 2 TV previews, cover of ‘Times 2’
	26 April 2013	1½	2	4		1 obituary; 1 TV preview

* all Fridays

† fractions indicate brief pieces

This approach enables a proportional observation to be made regarding the number of articles relative to the number of pages in the newspaper, as in Table 3:

Table 3: Samples of music content on Fridays in The Times relative to pagination.

		Number of music pieces	Total number of pages in newspaper	Proportion
1950s	6 April 1951	3¼	10	0.325
	13 July 1951	2	10	0.2
	5 September 1952	3	12	0.25
	24 April 1953	2¼	14	0.16
	11 December 1959	5	24	0.21
1980s	10 April 1981	4½	32	0.14
	17 July 1981	5	36	0.14
	10 September 1982	4	20	0.2
	29 April 1983	3	26	0.12
	14 December 1984	3	34	0.09
2010s	8 April 2011	13	120	0.11
	15 July 2011	13½	118	0.11
	7 September 2012	14½	132	0.11
	26 April 2013	9½	112	0.09

* all Fridays

This is a reasonable, if blunt, tool for a snapshot comparison between and within newspapers up to the latter decades of the twentieth century. On this basis, it would appear that the numerical increase in music articles and reviews is countered by a decrease in coverage as a proportion of the newspaper, notably between the 1950s and the 1980s. This is a scenario that is replicated for many subjects traditionally covered in newspapers. However, the raw figures are distorted by several factors. First, the relatively recent decrease in page size for most broadsheets means that more pages are required for the same amount of material.⁵ Second, the increasing ease with which photographs can be incorporated into the page has been a key factor in driving the expansion of newspapers. In other words, a significant proportion of the increase in overall size has come without a corresponding increase in text-content. Third, aside from this greater emphasis on image-based content – or possibly as a result – national broadsheet newspapers have added numerous sections, devoting much space to types of material, such as interior design, gardening, travel or cars, that hitherto was included only fitfully, if at all. In this respect, comparing a 1950s edition of The Times, when continuing post-War paper shortages restricted it to just ten pages, with today’s equivalent is like comparing a modest corner shop with an out-of-town modern supermarket. The two differ in so many respects that an unmediated direct comparison is flawed. There may be the same amount of coverage of music, or even an increase, but, relative to the whole, the proportion of space devoted to music (or many other subjects) would appear to have decreased markedly. All of this is without taking into account the lack of nuance in merely counting articles. Indeed, it raises a significant caveat about any quantitative approach. Namely that it does not necessarily follow that importance can be equated with quantity.

A relatively sophisticated example of a purely quantitative approach can be found in the article by Schmutz, Venrooij, Janssen and Verboord examining trends in coverage of popular music in ‘elite’ newspapers in the US, Germany, France and the Netherlands between 1955 and 2005.⁶ Examining two leading newspapers from each country, the broad methodology employed was to count and codify music content from one week per quarter as either ‘popular’ or ‘classical’, and also into one or more subgenres. In essence, the authors regard ‘popular’ music as being everything that is not ‘classical’, for they explicitly ‘use a broad definition of popular music as consisting of jazz, rock, R&B, blues, country, electronica, “pop” music (i.e. Top 40), world music, easy listening, brass band, and various regional genres such as chanson (France), schlager (Germany), and smartlap and kleinkunst (the Netherlands)’.⁷ This all makes sense for the analysis of broad trends, and is methodologically sound. In addition, there is a further quantitative element to the article, for ‘each article was measured in square centimeters, to provide a key indicator of the amount of space given to popular music by each newspaper’.⁸ As indicated above, such a figure would be relatively meaningless in isolation, as Schmutz et al appear to be aware: ‘To give an idea of the relative position of popular music, rather than of its absolute amount of space only, we provide comparisons with the amount of newspaper space occupied by classical music’.⁹ It is vitally important to remember that this article is written from the popular music perspective. It is attempting to assess the extent to which popular music has gained cultural legitimacy, and is written from the perspective of a relatively young discipline fighting to be taken seriously. The article explores possible reasons for the ‘lack of attention to popular music’ in German newspapers,¹⁰ and ‘the continued focus on classical music’.¹¹ However, examined through the other end of the telescope, the headline results of their research would appear to legitimise the anecdotal concerns raised by many in classical circles that popular music coverage has undermined, effectively stolen, space previously devoted to classical music:

In general, the findings point to the increasing legitimacy of popular music over time in all four countries. However, the size and timing of that shift varies substantially across countries. In particular, the US exhibits a much larger increase in popular music coverage relative to classical music coverage between 1955 and 1975, moving from 11.9% to 47.1% of the total space while none of the European countries allot more than 21.8% of the space to popular music in 1975. By 1995, however, the US, France, and the Netherlands all devote more newspaper space to popular music than to classical music. Meanwhile, Germany stands out in the relatively low amount of attention it gives to popular music, with classical music occupying more than twice as much space in the elite newspapers. Although the German newspapers pay a little more attention to popular music in 2005, it is still less than 40% of the overall space devoted to music; France still gives a majority of space to popular music (53.8%), and the US and the Netherlands give nearly two-thirds of the space to popular music (64.1% and 66.3%, respectively).¹²

These figures appear to suggest that classical music coverage has been substantially eroded – slashed, in the anecdotal vernacular – to make way for popular music, supporting a broader cultural perception of classical music being marginalised. While the latter may or may not be true, the figures in the article by Schmutz et al are seriously misleading without scrutinising their data in a way overlooked by the article. Moreover, from the classical perspective at least – and this ought to apply to the popular music perspective as well – the figures are founded on a flawed perspective that reflects a fundamental misapprehension of what has actually happened in newspapers over the past 50 years. The problem is that the article implicitly places popular and classical music as being in competition, opposition even. This easy fallacy reflects the anxieties and prejudices of both camps(or, rather, those who view themselves as belonging to a camp, regardless of which it is).

Just as it is of limited value to have quantitative figures for the number and size of articles of popular music on its own, percentage figures of popular music against classical music are not truly meaningful in isolation. What emerges from the sample figures for The Times is that classical music coverage has expanded, in numerical terms, but that popular music has appeared, then grown much faster. In the case of The Times (and the same is true of the Guardian), popular music coverage has been additional to that for classical music, not at the expense of it. In other words, coverage for music as a whole is significantly more, in straight numerical terms, than 50 or even 30 years ago. This is within a context of newspapers expanding substantially in the past 25 years ago, and takes no account of additional online content such as blogs, web discussions, exclusive articles and longer versions of articles and interviews.

Unlike Schmutz, Venrooij, Janssen, and Verboord’s study, the BritCrit Project deals nearly exclusively with digitised newspaper archives. It is important to be aware of the limitations of these digital resources. There is a huge amount of material available at a mouse-click. However, this has been scanned using OCR software, generally without subsequent proof-reading. While an image of the original is viewed, the search facilities rely on the accuracy of the software scanning. Due to the degradation of the originals, there is an increased likelihood of error the older the material. As a consequence, the search facilities for digital newspaper archives are far from reliable with results missing. For many types of research, this inevitable shortcoming either has no significant impact or is substantially outweighed by the general increase in access. However, for research that is attempting to chart trends across time periods, the knowledge that material is likely to be missing would undermine confidence in the absolute reliability of the findings.

A second caveat regarding digital archives is that they can distort perceptions of the material, favouring text-content, and giving all text equal status. In other words, the way that digital archives are accessed, which is primarily through a search facility, undermines visual factors. While this is significant for older newspapers, it is fundamental to newspapers from the latter half of the twentieth century and, especially, those from the late 1980s onwards. The increase of pictorial content, and the advent of digital typesetting, radically changed the presentation of newspapers. This is completely lost in the Nexis UK database, which provides only the text, with no indication of context, layout or related material such as photographs. It is essential to be aware of such potential pitfalls of digital newspaper archives when conducting searches. It is important to guard against developing an over-reliance upon the technology and to remain conscious that the results returned may not actually represent the sum total of all possible data, or be skewed by false positives or other errors. It is a result of these significant caveats about the digital archives that the sampling aspect of the BritCrit project mines material on the basis of a full reading of the chosen issues of the newspaper.

3. Working with The Times Digital Archive

Although a plethora of criticism has been produced in the second half of the twentieth century in the UK for the specialist (though non-academic) music magazines – such as Gramophone, BBC Music Magazine, NME, Q Magazine or The Wire – this project is specifically concerned with music criticism which has been produced for newspapers. Although it is a long-term ambition to extend the parameters of the investigation to other newspapers, the initial stages concentrated upon The Times with some consultation of the Manchester Guardian/Guardian. The Times is widely acknowledged for its extensive arts coverage and quality. Despite The Times’ long-standing commitment to classical-music coverage, the newspaper has never cultivated a particular editorial line. Thus individual critics were allowed complete artistic and critical freedom in the reviews which they produced.¹³ In this way, the criticism pages of The Times functioned as a discussion forum where critics could openly debate a whole range of musical issues. In working with criticism produced for The Times, therefore, a range of different opinions emerge, making it an ideal publication to work with, from a reception-studies perspective.

The digitisation of the newspaper, through a readily accessible portal, meanwhile, presents another real advantage.¹⁴ The Times Digital Archive presents a full-text facsimile archive of the newspaper from 1785 to 2006. Working with a digital resource overcomes the traditional drawbacks associated with archival research; namely, the necessity of conducting research which is both geographically and temporarily restricted (by the archive’s location and opening-times.) Digital archives represent a significant democratisation of knowledge via facilitating access to primary sources.

The extent of music criticism which has been published in The Times since the Second World War makes it impractical to survey it in its entirety. To provide a rough idea of the scale of music criticism produced for The Times since 1945: for the years 1945 to 1985 alone, one page per day was dedicated to arts criticism. Thus, one page per day, six days a week for a period of 40 years equates to over 12,400 newspaper pages which could potentially be analysed.¹⁵ For the initial stages of the BritCrit Project, therefore, the archive was surveyed at approximate five-year intervals instead: each providing a snapshot of the state of criticism at that time.¹⁶

This method of detailed examination of given points in time is familiar in the field of historical geography as ‘slices’. There, it allows ‘thick description’ of any one, or several, moments in time.¹⁷ When applied to criticism research, taking snapshots (or slices) of the state of criticism at five-year intervals allows us to build up a rough idea of how it is evolving over the greater period of time since the Second World War. Five-year snapshots were taken from two different times in the sampled years, from August, as a constant, and from a month chosen at random, as a variable. August was chosen as the constant due to the high number of international music festivals which traditionally take place during this month, including the Proms, the Salzburg Festival, the Edinburgh Festival, and the International Eisteddfod. All of these festivals have consistently attracted extensive coverage in The Times during the timeframe examined.¹⁸ Table 4 details the years and months from which criticism was initially collected:

Table 4: Months and years for which criticism was collected.

Year	August Dates	Randomly-Chosen Month Dates
1946	12-17 August	3-7 June19
1950	14-19 August	3-8 April
1955	15-20 August	4-9 July
1960	15-20 August	7-12 March 1960
1964	10-15 August	16-21 March
1970	17-22 August	4-9 May
1975	25-30 August	10-15 February
1980	18-23 August	6-11 October
1983	8-13 August	20-25 June 1983

All music-related articles were included amongst the criticism collected for the chosen weeks, encompassing reviews (both of recordings and of live performances), editorials, articles, interviews, and announcements.

The articles were initially subjected to qualitative textual analysis with a particular concentration upon the works and performers chosen for review, the locations of the performances reviewed, and changes in the language used within the reviews. Numerous changing trends were immediately apparent, notably the emergence of a new vocabulary to describe the early music performance movement, a pronounced shift in the criticism of female performers, and the existence of a much more eclectic concert life in post-war London than is often assumed. Whilst such qualitative analysis provides a firmer foundation for remarks on the changing nature, and current state of music criticism, the blended methodology adopted by the BritCrit Project allows such qualitative observations to be placed within a framework of quantitative evidence. In the first step towards quantitative research, precise metadata was recorded for each review collected. Many of the metadata fields which were directly related to the articles, including newspaper title, day of week (on which the article first appeared), calendar date, issue number, pagination, title of article, subtitle of article, author, layout, and cross-reference, were influenced by those used to reference articles by the Francophone Music Criticism Project.²⁰ Metadata fields were also compiled which related specifically to the works, composers, and performers being reviewed, including composers (discussed in the article), works, performers, venue, and date of performance, as the intention was to use statistics from the metadata to evaluate trends in the types of genres given critical attention over time. This metadata was recorded in tabular format, as exemplified in Table 5:

Table 5: Metadata example (1950 review of New York City Ballet).

Newspaper Title	The Times
Day of Week	Monday
Calendar Date	14/08/1950
Printed Date Correct	Yes
Issue Number	51769; Col21 C
Pagination	6
Title of Article	‘New York City Ballet’
Subtitle of Article	“The Four Temperaments”
Author	Anonymous
Layout	Internal main text
Cross-reference	None
Composers	Hindemith
Works	Four Temperaments
Performers	New York City Ballet, Choreographer: George Balanchine, Named dancers: Melissa Hayden and Francisco Moncion.
Venue	Covent Garden
Date of Performance	Saturday 12 August 1950
Brief Description	Review of a new ballet on music by Hindemith, depicting the four temperaments, choreographed by Balanchine. He used solos in Melancholy and Phlegm and Neo-classical procedures.
Images	None

Metadata collected from the reviews analysed was then inputted into a fully-searchable database (see Figure 1 for the Unified Modelling Language diagram).

Figure 1: Database Unified Modelling Language Diagram²²

The database was designed to be able to store the metadata in as detailed a form as possible; it is derived from the metadata table structure through an iterative normalisation process (Normalisation is a database re-structuring method that ensures that each table in the database contains only data that is very strongly linked. For example, pagination and word count belong to the article directly and are thus in the article table, while the name and gender of a performer might be mentioned in the article, but are attributes of the performer and are thus moved into their own table). Additional attributes of the metadata, such as the gender and nationality of composers and performers, are inputted into the database to enable searches which provide the maximum amount of information about the subjects chosen for review to be garnered. The central element in the database design is the article, which stores the core attributes of the article; linked to this are a number of tables that store information about performers, performances, musical works, composers and venues that are mentioned in the article. This structure enables very fine-grained searches in the quantitative analysis phase and also ensures that as little duplicate information as possible is stored. The advantages of a fully-searchable database of this kind are readily apparent. Most significantly it allows any claims about the changing nature of the criticism at The Times during the period covered by the corpora to be supported, or otherwise, by statistical evidence.

Applying this blended methodology facilitates the identification and charting of trends. In particular, it can help study patterns around canon formation, shifting attitudes towards women composers/performers, the reception of controversial or ‘problematic’ composers – such as Puccini – programming, and audience relations. Additionally, the results of this blended approach are often revisionist or counter-intuitive. For example, initial findings indicate that much more progressive, and diverse, attitudes to new music may have existed towards New Music in the UK in the 1950s than commonly believed. Both qualitative and quantitative aspects produce surprises.

4. Unexpected Outcomes

Exploiting the resources of digital newspaper archives enables a much more nuanced overview of how music criticism has changed over a substantial length of time than hitherto possible. One of the clearest trends to emerge is that music coverage in UK newspapers is variable. It always has been variable. Contrary to popular belief, music coverage has actually expanded in purely numerical terms. In fact, the golden age of music criticism may be the present or, at least, the past decade. Popular-music criticism has emerged as a distinct strand of music journalism. This, however, is in addition to, not replacing, coverage of classical music. These initial findings are at sharp variance with widely-held perceptions, to the extent that the current authors have even experienced angry responses in the face of the facts being presented.

Full exploitation of the blended methodology developed for this project provides a new tool for reception studies. Musicologists currently tend to use criticism as the raw material for understanding the reception of works and musicians, or to gain a greater understanding of the cultural contexts of music, its institutions and audiences. Criticism is rarely used as a revisionist tool to challenge the basic facts upon which the narratives of music history are constructed. The findings open the door to a more nuanced perspective of post-war music history that reflects the messier contemporaraneous reality of events more closely than the narratives of current histories. It is important to stress here, however, that the BritCrit Project has no a priori revisionist agenda; the expectation was that the findings would confirm the prevailing view of music criticism being in decline. It was unanticipated that the trends already uncovered –a more eclectic concert life in post-war London than articulated in existing narratives; the establishment of Boulez, Messiaen, and Stockhausen as totems for the radical avant grade in the UK much earlier than expected; more nuanced attitudes towards Soviet music and musicians than is generally portrayed – challenge aspects of the accepted understanding of music since 1945. Moreover, it provides a much deeper understanding of expected trends, such as the decrease in the use of gendered language to describe women musicians and their work or the increase in respectability given to popular-music genres. Crucially, it appears that common perceptions of the state of British music criticism itself are contradicted by the evidence thus far.

5. Conclusion: The Significance of a Blended Qualitative and Quantitative Approach

It is true that our research thus far can only be representative of one newspaper, the period we have concentrated upon, and the specific reviews which were examined for this project. Nonetheless, it is readily apparent that combining qualitative textual analysis with quantitative statistical evidence is a significant innovation in historical musicology and has already yielded valuable results. Whilst the examples given here only reflect the research thus completed, applications for this methodology are extensive. Virtually any claim about the nature of later twentieth-century criticism – or criticism from any period for that matter, if this approach were to be adopted – could be tested, and then supported or challenged through empirical results supported by qualitative analysis. It is necessary to re-iterate that this article is intended merely as an initial overview of this new methodology and its potential applications. While it is true that music criticism, like all journalism, is currently undergoing significant changes, a comparative assessment of these adjustments cannot be made on the basis of an erroneous view of the recent past. The initial findings of this project would appear to warrant a much more substantive investigation into the false memories of recent musical history.

1. ebrecht, Norman .‘Norman Lebrecht’s Free Thought’, broadcast on BBC Radio 3, 16^th October 2008. Available from http://www.bbc.co.uk/radio3/freethinking/2008/free-thought/norman-lebrecht.shtml (Accessed 28/06/2013).
2. Rothstein, Edward. ‘Criticism: Since 1945’, Grove Music Online. Oxford Music Online. http://www.oxfordmusiconline.com/subscriber/article/grove/music/40589pg3 (Accessed 10/01/2012).
3. Ibid.
4. ‘column-inch, n.’ OED Online. Oxford University Press. http://www.oed.com/view/Entry/36692?redirectedFrom=column-inch (Accessed 28/06/2013).
5. From 2004, The Times has only been available in tabloid format.
6. Vaughn Schmutz, Alex van Venrooij, Susanne Janssen, and Marc Verboord, ‘Change and Continuity in Newspaper Coverage of Popular Music since 1955: Evidence from the United States, France, Germany, and the Netherlands’, Popular Music and Society, Vol. 33, No. 4, (October 2010), 501–515.
7. Schmutz, et al, ‘Change and Continuity’. 504.
8. Schmutz, et al, ‘Change and Continuity’. 505.
9. Schmutz, et al, ‘Change and Continuity’. 505.
10. Schmutz, et al, ‘Change and Continuity’. 505.
11. Schmutz, et al, ‘Change and Continuity’. 507.
12. Schmutz, et al, ‘Change and Continuity’. 505.
13. It is necessary to acknowledge here that all music reviews published in The Times were anonymous until the mid-1960s.
14. See http://gale.cengage.co.uk/times.aspx/ though please note that access is via subscription only.
15. This figure, moreover, does not include the extensive coverage which appeared in the supplements in the later years of this period.
16. This more detailed process was separate from the sampling of content on Fridays used as illustration above.
17. The authors are grateful to Dr. Fiona Richards for drawing their attention to historico-geographical time ‘slices’.
18. A future study made in, say, 2040, would need to include similar popular music festivals, such as the Glastonbury Festival.
19. There was no Saturday edition of The Times on Saturday 8 June 1946 (presumably due to the paper shortages which were still afflicting Britain in the immediate post-war context).
20. The Francophone Music Criticism Project maintains a searchable, digitized repository of nineteenth-century French music and ballet reviews. It is managed by the joint directors of the former AHRC-funded Network in Francophone Music Criticism, 1789-1914, Professor Katharine Ellis (University of Bristol/former Director, Institute of Musical Research) and Professor Mark Everist (University of Southampton), and is hosted by the Institute of Musical Research, School of Advanced Study, University of London. See: http://music.sas.ac.uk/fmc.
21. Col: Column.
22. The authors are grateful to Dr Mark Hall for technical advice on database design and quantitative search techniques commonly used within computational linguistics.