IF YOU LIKE THIS YOU’LL ALSO LIKE… SOME THOUGHTS ON MUSICAL TAXONOMY, PART ONE

Musicians resist categorisation. The line ‘There’s only two kinds of music, good music and bad music’ is a commonplace. So is the epigram attributed to Louis Armstrong: “All music is folk music – I ain’t never heard a horse sing.” Neither is a very helpful thought when you walk into HMV looking for a record you’ll enjoy.

For lots of reasons it’s becoming more important to work out a sensible way of categorising music. One reason is the changing pattern of music-consumption and the shift to online buying of MP3s. There are also computer scientists trying to objectivise and computerise musical classification. Yet another reason is the new musical landscape – the ‘continuous present’ in which all the recorded music of the past 80 years is universally available. This creates the problem of too much choice. Despite this embarrassment of riches, it’s not clear that individual musical tastes are expanding; they may well be narrowing. The music industry does little to encouraging the expansion of taste, because it’s more comfortable with a neat segmentation of niche markets.
We are seeing the erosion and dilution of national and regional musical traditions and their replacement by a global mosaic of niche markets. All this is abetted by the general ignorance and confusion of music-lovers, the narrowing of musical choice in mainstream radio and television, and the inadequacy of musical coverage in the press and magazines.

It remains important for musicians to understand distinct genres and styles, whether they want to preserve and honour them or to regard them as colours on their musical palette. Musical progress tends to come from cross-fertilisation between traditions.

The problem

It becomes ever more difficult to decide where to place a particular record in the vast jigsaw-puzzle of musical ‘product’. For the critic the phrase ‘genre-busting’ is a term of praise; for the online record retailer it’s an expression that makes the heart sink.

Deejays and journalists are desperate to invent new names for slightly-different kinds of music. Here are a few descriptions culled from record reviews in The Sunday Times Culture Magazine, 3rd October 2010: “primitively-recorded mesh of big-mouthed Brit-beat vocals, snotty punk noise, sugar-rush power pop and oblique outsider strategies”… “jet-engine-in-a-basement sound”… “baroque chamber pop”… “uncompromising grime”… “disarmingly pastoral”… “post-Bartokian”… “dark psycho-comedy”… “pop-noir”… “skewed electronica”.

So what’s a musical genre, what’s a sub-genre and what’s just a cluster of weird anomalies?
Even when a genre is generally agreed to exist, there may be no general agreement about what to call it. For example, certain mid-sixties British pop groups (post-Merseybeat, pre-psychedelia) are said in continental Europe to belong in a category called ‘Freakbeat’ – a word that means nothing to most British music-lovers (although the term was invented by a British music journalist, Phil Smee).

The data

We have to start by asking what we can know about a record by listening to it. We can also consider the information that may be available on the label or the packaging and from external sources. Here are the main descriptors that are commonly used to differentiate examples of recorded music: Date – Place – Ethnicity – Precursors – Instrumentation – Production – Tempo – Structure – Mood and Message – Contexts of Performance – Contexts of Consumption – Markets and Marketing

Musical taxonomy is a multiverse with more dimensions than you’ll find in Stephen Hawking’s explanation of M-Theory. Each of the dimensions has its advantages and its drawbacks. For example: categorisation by date sounds straightforward, but music can be catalogued by the date of its composition, by the date of the particular performance in question, by the date of the record’s release (which may be years later) or by the date when it became successful and entered the charts (if indeed it ever did). The music may have been released on record at different times in different countries. It may be reissued, included in a compilation, used in a movie soundtrack or exploited by a TV commercial. It may appear in different formats, from 78rpm record through to 45rpm single, LP, cassette, cd and cd-dvd, and it is likely to be available in digital form, usually an MP3 or WAV file, as an internet download. Each new incarnation will have a date, but many of these dates will be undiscoverable without a lot of research.

Diagrammatic approaches

Biologists have used a hierarchical ‘tree of life’ structure to diagram the evolution of species in the natural world, and for them that works very well. The earliest attempts to construct this system were made in the 19th century. But the reason this approach works better for biology than it does for music is this: in nature, species do not interbreed and produce viable offspring. So whereas you can have lots of breeds of dogs (pedigree and mongrel), all descended from their common ancestor the wolf, you cannot (outside medieval mythology) have a cross between a horse and a crocodile.

If any species could breed with any other species to form a new line of development, biological taxonomy would get very messy. And yet that is what happens in music all the time. So we need a system of classification for music that allows us to describe and record all these matings and to follow the progress of their progeny.

Pete Frame

Pete Frame is a music journalist, best known for producing intricately-detailed outlines of the history of rock bands for various magazines. The family trees that he draws are a very useful way of tracing the careers of individual musicians and groups, but his method doesn’t translate very well into the wider sphere of styles and genres.

Closely related to the ‘family tree’ structure is the spider diagram, which seeks to plot the relationships between musical genres. Each style is encased in its own little box, and arrows connect them, each arrow indicating the influence exercised by another genre. A good example is this one, based on the diagram chalked up on the blackboard by Jack Black in the movie ‘School of Rock’.

There are many things wrong with this chart, but it offers an example of one way of mapping some of the territory.

Here’s another ‘family tree’ diagram that’s been published on the net. It’s designed to chart the development of UK pop music, but it betrays great ignorance of the subject.

Venn Diagrams

As the major problem of musical taxonomy is the problem of overlaps, we might find it helpful to express genre-relationships in Venn diagrams. For example if you draw three overlapping circles and you label them ‘Blues’, ‘Country’ and ‘Rhythm and Blues’ (by which I mean R&B from the 1940s onward, as distinct from ‘Modern R&B’) then you get four overlap areas, which might represent particular kinds of ‘fusion’ music .

This diagram can also be expressed as a set of equations:
country + blues = western swing
country + rhythm and blues = honky-tonk
blues + rhythm and blues = urban blues
country + blues + rhythm and blues = rock’n’roll

Sliding scales and distribution graphs

Another way to locate a piece of music is along a sliding scale between two genre-labels. So for example you can put jazz at one end and soul at the other end and you can string your records along it – Betty Carter towards the jazz end, Randy Crawford slightly further towards soul, Aretha quite a bit further along etc. Or you could equally well label the extremes ‘soul’ and ‘gospel’ or ‘jazz’ and ‘pop’.

For a more sophisticated version of this idea, we can combine two such sliding scales and create a distribution graph. Suppose we take the line that runs from “primitive folk music” to the most extreme varieties of “avant-garde art music” and make that our X axis. The let’s take the scale that runs from “rhythm-without-melody” at one extreme to “melody-without-rhythm” at the other. We can make that the Y axis. Now in theory we should be able to locate any piece of music / performer / album as a dot an the graph according to its distance from each of the extremes. A cluster of dots may then be labelled a genre.

Here’s an example of this sort of graph:

Another way to structure a diagram is to draw concentric circles and then divide each of the levels, from innermost to outermost, into segments. This gives us a reasonable chance of being able to plot where each genre touches its neighbours.

Here’s an example (from http://www.constantthoughts.net/wp-content/uploads/rock.gif)

I find this particular diagram deeply unconvincing.

In my next blog on this theme, I’ll put aside the theory and talk about how music is categorised in practice.

Advertisement

3 Comments

Comments RSS TrackBack Identifier URI