Bob Dylan has been recording in studios since the autumn of 1961. As of 2010 he has released 34 studio-recorded albums, 13 sets of live material and a further seven CDs of studio outtakes, with other ‘rarities’ sporadically spread across various compilations. Information and bootlegs circulating among fans suggests that there are many more high-quality unreleased studio recordings housed in the archives, featuring arrangements of songs written by Dylan and others.

The claims that are made about Dylan's contribution to the history of popular music are continually supported by reference to songs known through their studio recording. Yet, the activities of Dylan as a producer of recordings – his phonographic imagination – and his collaborations with those who have contributed to these productions have received little attention.

There are a number of reasons for the neglect of Dylan's sonic imaginings in the studio. Most obviously, the majority of critical writings about Dylan's art have concentrated on his song lyrics, usually treated as poetry on a page and analysed through approaches and assumptions drawn from literary criticism (Day Reference Day1988; Ricks Reference Ricks2003). When the lyrics have been acknowledged as the words of songs there has been some attention to vocal phrasing (Bowden Reference Bowden2001; Muir Reference Muir2003) but little recognition of the musical characteristics of these songs, and an avoidance of issues relating to production.

Dylan's studio work has also been overlooked because his most publicly audible music making (particularly during the early 1960s and since the late 1980s) has occurred on the stage. Paul Williams (Reference Williams1990, Reference Williams1992, Reference Williams2004) has been the most influential critic to argue that Dylan has always been primarily a ‘performing artist’. Williams has emphasised how Dylan has continually re-arranged songs in concert as a way of redefining or even undermining the studio recording as the definitive artefact. Such a perspective can be further maintained with reference to Dylan's reported comments about recording as a process, and his ambivalent and sometimes dismissive attitude to his recordings. Audio and visual recordings, along with published and internet-circulated concert reviews, have been regularly cited to support the claim that the most compelling renditions of songs have been in live performance, often considered expressively superior to the songs as recorded in the studio. It's an argument that I have incorporated into my own book about Dylan's music and musicianship (Negus Reference Negus2008). Yet it's a viewpoint that I have also begun to reassess.

My reassessment entails engaging with a further generalisation. This is the notion that Dylan's work in the studio has been largely spontaneous or accidental, a claim often extrapolated from a few anecdotes particularly, but by no means exclusively, alluding to mid-1960s sessions. These characteristically portray the producer's role as allowing Dylan to record with few obstructions, putting a microphone in front of him or arranging microphones in such a way as to follow his movements. In Colin Irwin's account of the making of the album Highway 61 Revisited, he quotes bass player Harvey Brooks: ‘Bob worked really spontaneously and fast and we didn't spend a lot of time looking for the perfect notes, it just had to feel right’ (Irwin Reference Irwin2007, p. 165). It is a typical comment. Dylan is assumed to be spontaneous because he is not concerned with correcting, overdubbing or editing notes. Apart from the fact that most music making is not about perfect notes anyway (this is surely a perception informed by the anxieties of session musicians), the ‘feel’ has to be worked for and created. Indeed, one of the reasons why Dylan has chosen (and dismissed) producers is for their ability to create a particular feel or soundworld.

Over the years Dylan has employed a number of recognised producers. For example, Jerry Wexler was asked to produce Slow Train Coming (1979), an album that was controversial due to the Christianity which dominates the lyrics and also commercially successful due to the tightly focused songs and sharp production which blends impassioned vocal performances, funky gospel rhythms and smooth blues licks from Mark Knopfler's guitar. Dylan chose Wexler for the sound he had created with Aretha Franklin and at Atlantic Records, and allowed Wexler (and his co-producer Barry Beckett) to create tightly arranged rhythm tracks prior to adding his vocals (Heylin Reference Heylin2000). Dylan then left the producer to overdub instruments along with harmony and antiphony vocals (whereas previously Dylan had preferred to record vocals simultaneously as the music was being recorded, with few later overdubs). Dylan also employed Don and David Was, who were responsible for the smooth (if bland) pop production on Under the Red Sky (again constructed using many overdubbed vocals and instruments); he worked with Daniel Lanois, who introduced his distinct multi-dimensional ambience and atmospheres to Oh Mercy! and Time Out of Mind, and brought in Arthur Baker to add a crisp, typically synthetic mid-1980s pop-dance sonority to Empire Burlesque. After hearing the first mixed and completed version, it was Baker who asked Dylan to write an additional acoustic song to close the album, and ‘Dark Eyes’ was composed on the night it was suggested and added as the final track (see Dylan Reference Dylan2005, pp. 209–10).

Dylan's decision to work with these (and other important producers), and the audible evidence of their input and of his many albums having a very distinctive overall sound – often in contrast to preceding and following albums – suggests to me that Dylan has entered studios with a very clear purpose, his work guided by clear sonic intentions, even if the songs have initially been ‘unfinished’ and the final musical arrangement and full text of lyrics completed during the process of recording. The fact that Dylan has frequently rewritten songs in the studio as arrangements have coalesced and suggested new angles is itself an indication of the importance of the studio in shaping and extending his songwriting. Despite all this, most commentaries on Dylan seek to downplay the significance of producers and studio production throughout his entire career. Tom Wilson is often quoted: ‘You don't think of orthodox recording techniques when you're dealing with him. You have to learn to be as free on this side of the glass as he is out there’ (Irwin Reference Irwin2007, p. 113). This is a fairly typical remark, again from the 1960s, used to buttress the argument that the role of producers has been to capture a fresh performance in a way that allows Dylan to be as relaxed as possible. There is clearly an issue here. A number of reports portray Dylan as rapidly becoming irritated, bored and frustrated when recording and rehearsing with other musicians. However, such a scenario all too easily fits the image of Dylan as mercurial, intuitive and awkward. It obscures the preparation, critical reflection and deliberation that have informed his studio recordings. It downplays the considerable effort that Dylan has put into production when attempting to find the most suitable version of a song – even if that recorded version will then be subject to re-arrangement on stage as Dylan's musical ideas change.

Dylan's frustrations with the studio have been shaped by a very specific era in the history of recording. He has lived through the introduction of multi-track technology and observed the studio become a customised (even clinical) professionalised environment with a structured emphasis on accuracy and repeatability. From a prior emphasis on capturing the dynamic performance of an ensemble, the studio became a space where musicians (often playing alone) lay down multiple takes of songs which can be edited, overdubbed and mixed to arrive at the definitive version of a song, with an imperative to record different takes in an identical way (the same tempo, key, timbre, instrumentation) so that they can be edited together. Increasingly, machinery has been used to regularise beats and tunings, feeding a culture of technological perfection that avoids or erases any evidence of ‘mistakes’ on released recordings. Recording (whether a rock band or orchestra) has become the means of arriving at the ideal arrangement and ‘performance’ of a song or symphony.

While Dylan has challenged the ethos of perfectionism inherent in these aesthetic values and technological practices, he has also selectively employed them as he has used the studio, often in subtle unacknowledged ways, as an integral part of the process of composition – as a means through which the possibilities of songs are explored and realised. This is one aspect of Dylan's music and musicianship that seems most concealed in critical writings and my overall point in this article is to argue that Dylan's studio practices should be subject to as much sustained research, analysis and discussion as the effort expended in explicating his wordplay. He has produced his last three albums of new songs – Love and Theft (2001), Modern Times (2006) and Together Through Life (2009) – using the pseudonym of Jack Frost, and adopted a very clear phonographic aesthetic, one which he has touched upon directly or obliquely during interviews, particularly when praising the sound on old Chess and Sun records (a sonority, incidentally, that featured prominently in his ‘Theme Time’ Radio Series).

In using the phrase ‘phonographic aesthetic’, I am aware that I am introducing a word (phonography) that has been adopted in specific, but varying, conceptual terms in musical scholarship (used when referring to field recording, or composition using ‘disembodied’ and abstracted sounds, or a distinctively modern understanding and experience of sound).Footnote ¹ The term has also been used in language and linguistics (in discussions of systems of phonetic spelling and writing) and also bandied about in a more casual manner (when it has appeared as the title of an album by R Stevie Moore and a track performed by Britney Spears). My use of the term is most clearly indebted to its adoption by Evan Eisenberg (Reference Eisenberg1988) when questioning assumptions that sound recordings can be heard as a ‘record’ (or ‘sound photograph’ – a commonly used analogy) of a performance that exists (or existed) independently of the recording. Eisenberg's argument is that a record (sound recording) should be understood according to its inherent characteristics rather than in relation to a ‘live’ event. Phonography implies a performance that is constituted or realised in the very act of recording; the ‘art of phonography’ is the ‘composite’ construction of ‘an ideal event’ (p. 89). In some respects this might be thought of as a continuum; at one end could be a recording that is ‘sketched’ quite rapidly through a more or less ‘live’ solo or ensemble performance in the studio; at the other end a carefully assembled sound collage, built up from assorted elements, a combination of many recordings created over days, weeks, months or years, and mixed and remixed to create different potentially infinite versions. The received wisdom in much writing is that Dylan's recordings occupy the first point on this continuum, having usually been laid down quickly.

There seems no reason to unduly challenge accounts of the time Dylan has spent in the studio; this can be reasonably substantiated. Yet, to think about Dylan's phonographic imagination implies a more profound process, most notably in the way that recording allows for the radical separation of a performance from its contexts of ‘origin’ along with the abstraction of that moment of social, collective or public musical creation into a more private, subjective and solitary space (Eisenberg Reference Eisenberg1988). Recordings can be removed, taken away and subjected to intense, repeated listening in a way that is in marked contrast to the experience and then recollection of any performance (as those who have listened to bootleg recordings of remembered concerts will frequently attest). In his discussion of the recording of Oh Mercy, Dylan frequently mentions this, at one point recalling that he ‘took a tape of the song with me to study and headed back to the house … Later that night, listening to what we'd done, I thought I'd figured it out’ (2004, p. 183). There are a number of accounts of Dylan re-recording songs (and wishing he could re-record songs) after having listened to the work created in the studio; one of the most celebrated examples was his dissatisfaction with some takes produced in New York City for Blood on the Tracks, and his subsequent decision to re-arrange and re-record a number of the songs in Minneapolis (see Gill and Odegard Reference Gill and Odegard2004; Negus Reference Negus2008, pp. 47–48). Hence, it is not just the time that Dylan has spent in the studio that is an issue. Also neglected has been evidence of the preparation and reflection that has gone into those moments and processes in the studio, and the way Dylan has devoted considerably more effort to the post-production period of mixing and remixing, and sometimes re-recording, than is recognised.

As a way of exploring some of the significant qualities that have characterised Dylan's recording aesthetic I am going to identify four phonographic practices. I present these as just one path into and through this issue. It is a tentative route, allowing me to make a number of points on the way, and there is clearly much more to be researched and written about these aspects than I have the space to pursue here. First, I want to emphasise a recording aesthetic that Dylan has drawn from the acoustic blues and folk ballad which values irregularity in pulse and phrasing. While it is clear that the blues and folk ballad have informed Dylan's performances since the earliest days, here I want to stress the important influence of folk ballads and blues as recorded arts (not just as song forms or generic styles). Second, I will highlight the enduring influence of a sonority indebted to electric blues as it blurred and merged into 1950s r'n'b and rock'n'roll recordings. This is apparent in abrasive instrumental timbres, and a distinct approach to the balancing and placing of instruments in the mix. The third is an approach to recording textures that is indebted to 1960s pop productions, particularly those of Phil Spector – not by any means imitative of Spector, but influenced by the way Spector strove for an overall sound or texture, a blend rather than a careful separation of instruments. Fourthly, I want to draw attention to Dylan's cinematic approach to recording songs. In delineating these contours of Dylan's phonographic imagination, I do not propose to account for nearly 50 years of recordings, but rather to distinguish significant threads that are dominant at certain points and woven throughout Dylan's recording history.

My first point, concerning the legacy of Dylan's early immersion in the blues and folk ballad traditions, can lead some listeners to mistakenly infer Dylan's casual attitude to the studio, with the perceived imperfections of recordings cited as evidence. Eyolf Østrem makes this presumption in an often-insightful study that uses formal musical analysis as a means of understanding Dylan's repertoire. Of the album Another Side of Bob Dylan he writes: ‘In several of the songs … it is evident that Dylan hasn't really learnt the chord changes properly before he started recording… . it is difficult to find two verses that are played in the same way. There are lots of temporary solutions’ (Østrem Reference Østrem2008, p. 541).

I doubt that Dylan had not thought about how he wanted to record songs; it is improbable that he had not ‘learnt’ the chords to his own compositions. It seems far more plausible to hear this as evidence of the way Dylan has actively embraced an aesthetic of irregularity, an orientation that he absorbed from folk ballads and the blues. It can be heard on numerous recordings when Dylan is alone on acoustic guitar and harmonica, or piano. He will introduce subtle variations in tempo (slowing down, speeding up, seeming to hesitate). He will add additional beats, drop beats and appear to start a verse or chorus slightly early, and sing verses and choruses in slightly different ways. This is apparent throughout his early albums and noticeable on the album singled out by Østrem, Another Side of Bob Dylan, notable on ‘My Back Pages’, ‘Ballad in Plain D’ and ‘Chimes of Freedom’.

Rather than errors, or evidence that he hasn't learned the songs, it is indicative of a valued aesthetic quality he has absorbed from an immersion in listening to performances and recordings of folk ballads and acoustic blues, particularly the Harry Smith Anthology of American Folk Music which had a profound impact upon Dylan's sound and imagery (Marcus Reference Marcus1997; van Ronk Reference Van Ronk2006). The influence of the acoustic ‘Delta’ blues and folk ballad have been discussed extensively in Dylan scholarship, and this particular influence has been pursued by Charles Ford who produced a detailed study of irregularity in Robert Johnson's recorded performances (Ford Reference Ford1998). Ford made a similar study (yet unpublished) of irregularities in Dylan's songs up to and including ‘Subterranean Homesick Blues’. Of 124 songs, he found eight to be completely irregular, 81 slightly irregular and 35 irregular (personal communication). It is tricky catching these irregular inflections on later recordings when Dylan is performing with a band, as the arrangement and drummer's timing tend to regularise the pulse. Yet the irregularities are often still there. They are very noticeable on New Morning and on Oh Mercy, when Dylan is directing arrangements from the piano, and crop up in varying ways on later albums (Love and Theft, Modern Times and Together Through Life – all produced by Dylan as Jack Frost, all featuring members of his touring band being directed by Dylan largely playing ‘live’ as an ensemble for the recording).

It is commonly known that the blues had a profound impact on Dylan's songwriting, a fact he has regularly acknowledged. Not only did he absorb the aesthetic of irregular rhythms, of an ‘imperfect’ pulse (measured against the studio techniques I referred to earlier), he also acquired a method for constructing songs and a particular approach to timbre. In Chronicles Volume One, Dylan recalls how Robert Johnson made his ‘hair stand up’ (2005, p. 282) and explains how he spent weeks listening repeatedly to the recordings, infatuated with the sounds Johnson extracted from the guitar, with his evocative voice and his rhythmic sophistication. Dylan was equally obsessed with the narrative depth and metaphorical intensity that Johnson conveyed economically in his lyrics. Peter Guralnick (Reference Guralnick1989) and Greil Marcus (Reference Marcus1977) have both drawn attention to the way Robert Johnson's lyrics were qualitatively different to those of many other blues performers in the way he used metaphor and pursued narratives. Like so many popular musicians, Dylan learnt the blues, and acquired his techniques and his musical identity from repeated listening to recordings. Dylan's approach to performance, and to the studio, was shaped by an immersion in the records of blues musicians who themselves were notable for having acquired a ‘recording consciousness’, a term now used by many writers and initially introduced by H. Stith Bennett (Reference Bennett1980) in a study of how popular musicians acquire their craft skills and orient themselves to peers and performances through recordings that are subject to a process of repeated and intensive listening.

Eric Rothenbuhler has developed this theme when arguing that what we know of Johnson's music exhibits a deliberate ‘for-the-record aesthetic’ in which Johnson valued and recognised the potential of recordings; he understood that they had to stand up to frequent listening. Rothenbuhler (Reference Rothenbuhler2007) argues that Johnson did not simply record as he might have performed in a bar, entertaining an audience, but composed and structured his songs with the recording in mind.

Many writers have illustrated how recordings became central to blues culture. Marybeth Hamilton is just one researcher who has noted that during the early 1920s and ‘seemingly overnight, an oral culture was inundated with mechanical reproductions, as people long celebrated for spontaneous singing immersed themselves in the sound of recorded discs’ (Hamilton Reference Hamilton2007, p. 11). Although little is known of Johnson's life, there is evidence (and perhaps plausible speculation) that he was familiar with many styles and songs of the day, and that records – listened to intently and heard more ambiently from jukeboxes – influenced his approach. Rothenbuhler stresses how recording, the art of phonography, shaped the craft of Johnson's songwriting. It is likely that Johnson would have been equally aware of the sonority and timbre; how the sound of music was experienced from a recording – what Dylan heard as the ‘stabbing sounds from the guitar’ that ‘could almost break a window’, the singing voice that could ‘have sprung from the head of Zeus in full armour’ (Dylan Reference Dylan2005, p. 282).

If the acoustic blues, along with the folk ballad, was the most audible influence on the recordings that Dylan publicly released up to 1965, the electric blues, r'n'b and rock'n'roll (the boundaries blurred, the categories ineluctably reductionist) had the most profound impact on the timbre, texture and colour of sound that Dylan would attempt to create when he began recording with ensembles.

When Dylan first started recording with a band of musicians, it was with men influenced by the electric blues – its distinct sonority, its ‘mathematical’ structures (a term used by both Chuck Berry and Dylan), its lyrical patterns and its wider social and cultural significance at a time of racial antagonism and the struggle for civil rights. Steve Waksman, writing of the period from the 1950s into the 1960s and discussing how blues informed rock music, has argued:

Whether or not one accepts all of the points in this quote, when Dylan moved from acoustic performances and recording, and ‘went electric’ (as the cliché goes) he was not in any straightforward way rejecting folk principles for commodified mass culture (as was sometimes, and occasionally still is, asserted). He was incorporating an electric guitar sound as symbolic of black culture; the aforementioned Mike Bloomfield became central to Dylan's recorded sound on Highway 61 Revisited. If some of his protest songs were challenging racial divisions at the level of lyrics (delivered within folk ballads that were often construed as a white form, no matter how misleading this assumption and even if a clear black presence had become camouflaged), with a band he confronted this antagonism at the level of sonority (albeit with the contradictions that Waksman alludes to in the above quote).

Blues musicians had begun to use the electric guitar ‘for the ways in which it contributed to the overall sound’ (Waksman Reference Waksman1999, p. 147), not just as a lead instrument. Amplification extended the range of the guitar and its potential contribution to a band's sonic vocabulary. Waksman explains how amplification had an impact on a blues sound that would later shape Dylan's recording aesthetic. Blues musicians playing in clubs, seeking to gain and to hold their audiences' attention, developed an ‘aesthetic preference for sounds that cut against the grain of a smooth musical surface’ (p. 7) … ‘timbres that contrast rather than blend’ (p. 119). Regardless of whether the impetus for these sounds was indebted to African retentions or a specifically African-American impulse (Waksman cites some rather essentialist sources here), throughout the 1950s ‘blues guitarists would employ increasingly coarse, distorted tones, extending the expressive palette of the electric guitar in new directions’ (pp. 137–38). At the same time, and closely related to the way amplification was used to create additional tones, producers and engineers at Sun and Chess began to challenge recording conventions by allowing the dials to move ‘into the red’ (technically indicating ‘distortion’).

It is this abrasive timbre that characterises so many Dylan albums. In early recordings performed with acoustic guitar it is present in his voice. As his voice mellows, with closer mic-ing techniques on some of the mid-1960s recordings (notably Blonde on Blonde) the abrasive timbre transfers to the guitars and Dylan's equally harsh blues-inspired harmonica techniques.

This blues sonority became 1950s rock'n'roll, a music defined through its recordings. Rock'n'roll emerged from the continual interplay between amplification and recording. Recording did not follow, but informed the r'n'b/ rock'n'roll aesthetic. Chuck Berry, for example, developed his guitar playing ‘with a tape recorder at his side’ (Waksman Reference Waksman1999, p. 149). Like many musicians, Berry's exceptional and innovative style, his sonic identity was created according to how it sounded when played back on magnetic tape and then on disc. The act of recording and listening back shaped the way he used the guitar and structured his songs. Chuck Berry was a major influence on Dylan's mid-1960s band recordings, nowhere more so than on ‘Subterranean Homesick Blues’ which, as Dylan later acknowledged, is lyrically and musically indebted to Chuck Berry's ‘Too Much Monkey Business’. Using derivative rhythms and the same blues chord patterns, and recasting Berry's dissatisfaction with schooling, menial jobs, domesticity and respectability, ‘Subterranean Homesick Blues’ is a list of caustic observations (knowingly drawing from beat poetry) about the impact of institutions, authority and respectability.

Hence, not only did Dylan draw on the blues in his lyrical imagery and song structures (the pervasive use of what is known as the three chord 12 bar blues), the abrasive sonority of electric guitar blues shaped his recording aesthetic. In an interview with Bill Flanagan issued to accompany the release of Together Through Life, Dylan commented, ‘the old Chess records, the Sun records. I think that's my favourite sound for a record … I like the mood of those records – the intensity. The sound is uncluttered. There's power and suspense. The whole vibration feels like it's coming from inside your mind. It's alive. It's right there’ (Flanagan Reference Flanagan2009). The mood and intensity of those records has been on Dylan's mind for some time (presumably since the days when he first heard them), and has informed his approach to the studio since he began recording with an ensemble. It can be heard as a phantom presence even when he's experimenting with more cluttered textures (the harsh blend of violin and harmonica on Desire) or embracing more contemporary pop production techniques, as he did briefly in the 1980s, particularly with the metallic, synth-drenched timbres on Empire Burlesque (1985).

The phantom presence of Chess and Sun informed the production of Time Out of Mind, released in 1997, often cited as the album that marked Dylan's critical re-evaluation and his artistic re-emergence from a period of uncertainty on the recording front (Marshall Reference Marshall2007), and the last album to be co-produced. At the time, Dylan expressed his dissatisfaction with the quality of contemporary recordings, and records made during the 1950s informed its production. As I have discussed in more detail elsewhere (Negus Reference Negus2008) the musicians were initially recorded playing along with old recordings when working on arrangements, and samples from this were used in the mix. During the sessions musicians were physically located and microphones were strategically placed so as to produce a sonic perspective where the drums are often far back in the mix, behind the other instruments.

However, unlike 1950s records, Lanois filled the gaps with drones, percussion, sustained guitar or organ chords. Time Out of Mind was engineered to emphasise mid-range frequencies, with little at the top and bass end (again attempting to approximate a 1950s sound). Throughout much of the album a very quick single ‘slap back’ echo was added to the voice – a direct reference to the 1950s recordings of Elvis Presley. The production on the album was subsequently described as ‘swampy’ by Dylan and by critics picking up on this apt description when finding Lanois culpable of creating an unsuitable sound for Dylan's songs. The tracks were certainly not ‘uncluttered’. In fact, from the mid-1970s to the mid-1980s, Dylan had experimented with a variety of always intriguing (though not always critically appreciated) ‘cluttered’ and dense production textures, adopting a multi-layered approach to instrumentation and rhythms, exploring varied tones and shadings, most notably on Desire, Street Legal, Empire Burlesque and Down in the Groove. Listening to these alongside what Dylan referred to as the bright ‘wild mercury sound’ of Blonde on Blonde and the shimmering acoustic guitar washes of Blood on the Tracks provides aural commentary on how, along with his constant use of abrasive sonorities, Dylan has also been concerned with how instruments combine when creating individual albums with distinctive timbres and textures.

The attention to an overall sound and a blues sonority pervaded one of Dylan's most critically acclaimed (and lyrically analysed) albums, Highway 61 Revisited (1965). The rich musical texture not only drew from the blues and r'n'b, but also was indebted, whether consciously or not, to the dense ‘wall of sound’ created by producer Phil Spector during the early 1960s. As is well known, Spector had used larger than conventional studio ensembles, lined up musicians playing in unison, combined electric and acoustic instruments, and fed reverberation back into the recording as it was taking place. He immersed and buried individual instruments in an overall texture rather than allowing them to stand out with clarity.

Prior to recording this album, Dylan had disagreed with Tom Wilson, who had produced his previous three albums, and had suggested Spector as a possible producer. Many critics seem bemused by this. Mark Polizzotti (Reference Polizzotti2006, p. 78) assumes that Dylan was being ‘sarcastic’. Clinton Heylin (Reference Heylin2000, p. 217) worries that ‘what Phil Spector would have made of Highway 61 I fear to speculate’. Colin Irwin, equally concerned, writes ‘the mind boggles at what sort of album Highway 61 Revisited might have become had Spector been installed behind the controls’ (Irwin Reference Irwin2007, p. 114). This scepticism seems misplaced. Not long after the album was released, Dylan and Spector became acquainted and spent some time discussing music together (Shelton Reference Shelton1986). They clearly respected each other's work. Spector went on to produce the best solo work of John Lennon (the singles ‘Instant Karma’, ‘Power To The People’, the albums Plastic Ono Band, Imagine) and George Harrison (All Things Must Pass). Latterly, the production on Bruce Springsteen's Born to Run and Born In The USA is hugely indebted to both Spector's wall of sound and Dylan's mid-1960s instrumental textures. There is more than a hint of Spector's recording aesthetic on Highway 61 Revisited.

I would not wish to argue that this relationship can be demonstrated by playing extracts from songs produced by Spector alongside selections from Highway 61 Revisited. The connections are not straightforward – and I don't mean to imply a causal link between Spector's approach to production and Dylan's sounds. After all, Spector did not produce John Lennon and make him sound like early 1960s girl groups or the Righteous Brothers. Instead, he brought out important elements in Lennon's sound and the tendencies in his songwriting. He did not impose his dense layering techniques on Lennon, but sympathetically supported Lennon's sound – using space and echo to create atmosphere, and suggesting the adoption of favoured methods such as using many pianos and removing cymbals from the drum kit on ‘Instant Karma’, for example. Of Lennon's Plastic Ono Band album, Mick Brown has written of how Spector ‘fashioned a stark and spare production that perfectly matched the tenor of Lennon's songs’ (Brown Reference Brown2008, p. 260). Of Imagine, the same writer has also observed:

It seems plausible to suppose that Spector would have acted in a similar way as a facilitator of Dylan's sonic imaginings and I would argue that the production on Highway 61 Revisited can be taken as just one example of how musicians were influenced by Spector in realising their own distinct sound. Although Columbia employee Bob Johnston is credited with the production of this collection, Dylan actively directed the recording and mixing of the sessions. The sound is less relentless than Spector's wall, more a sea of sound with continual waves, in which the lyrics float in and out as passing images rather than stanzas of verse and in which the inspired musical gestures of individual musicians frequently rise to the surface: Mike Bloomfield's searing, metallic electric guitar; Paul Griffin's melodic embellishments on piano; Al Kooper's swirling organ; Dylan's characteristic high-pitched, insistent harmonica. But it is the overall dense sonic texture that so often strikes listeners when they first encounter the album, what Polizzotti hears as ‘something thick, enveloping, and elemental as mud’ (2006, p. 19). Recorded in the same year (1965) ‘Can You Please Crawl Out Your Window’ may also include an oblique nod to Spector in the use of a glockenspiel.

Highway 61 Revisited includes the much celebrated and analysed ‘Ballad of a Thin Man’ which blends Dylan's sarcastic commentary directed at the bemused Mr Jones with imagery drawn from the singer's memories of seeing geeks and freaks at carnivals during the 1950s (inspiration for the song Dylan recounted on stage at Charlotte Coliseum, 19 December 1978). The track is led by Dylan's stalking, accusatory piano; it is pervaded by spooky Wurlitzer-like organ flourishes straight out of suspense or horror movies and punctuated by touches of trebly and twanging guitar reminiscent of westerns and mysteries. It is as if the band is creating a soundtrack to a movie, and not simply an arrangement of a song (as they are on many of the tracks on this album). There are striking sonic parallels between ‘Ballad of a Thin Man’ and various horror film and television themes from this period. For example, listen to ‘Ballad of a Thin Man’ with Henry Mancini's edgy soundtrack to Blake Edwards' Experiment in Terror from 1962.

Wilfrid Mellers (Reference Mellers1984) and Robert Shelton (Reference Shelton1986) are just two writers who have emphasised the way Dylan's lyric writing has been informed by a ‘cinematic imagination’ (it may also be the case that Dylan's songwriting has drawn from novelists whose writing, since the 1930s, has been influenced by film). The assumption seems to be that Dylan's lyrical constructions are indebted to cinema in the poetic invocation of visual imagery and the presentation of narrative through episodic-like or montage-like verses (although this quality of Dylan's songwriting may be equally indebted to the ‘non-linear’ way that stories and point of view are conveyed in many old ballads). Yet, I'd argue that the influence of cinema is not only in the lyrics, but also in the musical arrangements and production.Footnote ² There are a number of examples of a cinematic approach to the production of tracks, particularly in those songs that unfold with a film-influenced lyrical narrative. An obvious track is the epic 11-minute ‘Brownsville Girl’, co-written with Sam Shepard from the album Knocked out Loaded (1986), an episodic or montage-like story that refers to an unnamed Gregory Peck movie in the lyrics. The track has a dense wall-of-sound production and musically signals and plays with melodramatic pop song clichés and Hollywood film score codes. The dynamics of the production – moving through dense swirling crescendos and falling to quieter, emptier passages – support Dylan's laconic, semi-spoken ‘voice over’ delivery and allow the song to accumulate a dramatic intensity through some dynamic interplay with female vocalists, as chorus and individual voices.

Another example of a cinematic aesthetic – of the track as soundtrack – can be heard on ‘Man in the Long Black Coat’, composed during the Oh Mercy sessions following ‘nagging’ from producer Lanois, a song Dylan acknowledged, ‘I never would have written otherwise’ (2005, p. 209). Fading in with the sound of crickets, the production uses echo and delay on instruments to evoke a gothic atmosphere of small town paranoia, with intimations of supernatural presences and grotesque distortions of ordinary life. Dylan seems to peer out from the shadows as a wailing harmonica sound echoes towards the edge of town, down deserted darkened streets. A clipped, stalking unequalised bass underpins Dylan's hesitant, anxious vocal delivery. Seeming to be gasping for breath, placing accents on unexpected syllables, his singing encapsulates the track's mood of strangeness and unease.

This is a phonographic, cinematic soundworld, a ‘for-the-record’ aesthetic constructed in the studio; both tracks make extensive use of echo and weaving in and out of instrumental layers when coding the emotional atmosphere and narrative direction of the songs. Dylan had grown up hearing echo and its associations on both records and film soundtracks. As mentioned earlier, he has often singled out the sound of Chess and Sun records as an influence; both labels pioneered the use of echo and reverb techniques during the 1950s. Dylan was a keen film fan when in his teens (Shelton Reference Shelton1986) and would have been familiar with the way echo and reverb were incorporated into film soundtracks from the 1940s and used to convey a sense of twilight, a darkening of space, a dimming of light, or to give the impression of things losing their solidity as they blend into the surroundings, and to conjure the presence of spirits, phantoms, mysteries or evoke a dream-like state (Doyle Reference Doyle2004). He would also have heard many blues and country records on which ‘echo and reverb effects were … used to suggest shadowy, subterranean, marginal presences’ (Doyle Reference Doyle2004, p. 39). Such sounds and their associations, absorbed through repeated exposure to recordings and movies, has had an audible impact on Dylan's phonographic imagination; it can be heard directly informing the production of his songs.

This point returns me to one of my main threads: Dylan's musical identity is not something that appears to listeners independent of its realisation through recording. Studio practice and the art of phonography are integral to Dylan's identity and songwriting. The process and practice of songwriting is integrally tied up with its realisation through the act of recording and the activity of listening to songs and performances from recordings, regardless of how much time has actually been spent in the studio.

In writings about Dylan it is frequently claimed that the studio is where songs are recorded quickly as spontaneous drafts that are later developed on stage, the live concert being the public arena within which Dylan resolves the enduring tensions between songs as multiple performances and songs as definitive recordings. From this perspective, his songs come alive on stage and challenge the frozen studio record. Yet the recordings endure and can be listened to time and time again. They do so because they are far from tentative or spontaneous drafts. Dylan may not have recorded in conventional ways. He may have appeared casual and sometimes reluctant. But his recordings are informed by distinct cultural and musical values, clear aesthetic principles and sonic intentions. Dylan's art is characterised by a phonographic imagination astutely attuned to the ‘recording consciousness’ of his listeners and peers, acquired from immersion in a rich history of recorded popular song and an awareness that each recording must offer something more – not just the same old Bob Dylan sound. A sense of his frustration with the process and the effort devoted to finding the right version is captured in the following memory, a tale undoubtedly familiar to many who have spent time in studios, again from recording Oh Mercy with Lanois:

In focusing on the legacies of the recorded blues and folk ballad, the influence of pop productions that are concerned with an overall sound, and a tendency towards visualising sound in a cinematic manner, I have suggested that Dylan devotes more care, preparation and thought to the recording process than is conventionally assumed. His memoir, Chronicles Volume One, is littered with judgements and observations about the ambience of studios and the influence of the cities within which they are located, the contributions of particular producers and musicians and even the placing of instrumentalists and microphones within the recording space. My overall point is that recording should be integrated into discussions of his art, alongside the attention devoted to lyrics, performance and star-image or public identity. In this essay I have provided just one route into these issues. There is much more to be researched and written about Bob Dylan as recording artist, and many signposts to further routes can be found in existing studies. We could, for example, focus in more detail on how individual songs are constructed in the studio. Greil Marcus' book-length study of Like a Rolling Stone begins to do this. Marcus celebrates the character and quality of this recording for the way it allows the listener to continually hear something new: ‘the sound is so rich the song never plays the same way twice’ (Marcus Reference Marcus2005, p. 97). Of the song in concert, Marcus is explicitly opposed to writers such as Paul Williams, and hears ‘a warhorse, trotted out one more time to circle the track … Often the music thinned, and the song was like a bad bluegrass tune … There have been a thousand performances where the same nothing happens, with more flair or less, but no real difference’ (p. 186).

Here the stark tension between studio recording and live performance is turned on its head. Marcus finds value in the special recording, the unique frozen moment and argues that it can continually provide new experiences for the listener. From this perspective it is the recording that illustrates the unfinished character of the pop song, due to the way we can appropriate it, absorb it into our changing circumstances and continually hear it in new ways. In contrast, the live performance is little more than an obligatory, nightly ritual, the familiar rendition merely reinforcing the authority of the recording.

As Marcus explains, drawing together insights from a variety of sources, the track had started life as a long rage-driven poem, Dylan saying that it had been ‘vomited out’ and that he wasn't too sure what to do with it. He subsequently began doodling around at the piano and started to develop a song. The early versions, even those tape-recorded the day before the well known recording was made, are characterised by a less declamatory, more uncertain ballad style. A number of writers have suggested that it is in waltz time, and Marcus even quotes from a recording tape from the session where producer Tom Wilson comments ‘It's a waltz, man’ and Dylan fires back ‘It's not a waltz’. (p. 1) Dylan is right. It most certainly is not a waltz. The fact that it has triplets or is based on a pulse that can be heard as divisions of three rather than four doesn't mean it's a waltz. To my ears it's a lilting 6/8 rhythm – you can pick out the slightly irregular triple time, but across this you can hear the 1–2, 1–2 swing – a characteristic of folk ballads and many of Dylan's earlier songs.

The early version that Dylan plays on the piano is in the key of C#/ D♭. This can be heard on the take made available on the second volume in the bootleg series and evidence of this can be seen in photographs taken on the first day of recording (reproduced in Marcus's book). Seated at the piano, Dylan clearly has his fingers on the black note chord shapes of such a key. In addition, Greil Marcus (Reference Marcus2005) has provided transcripts of some of the discussions at the session and reports Mike Bloomfield giving instructions to the band: ‘OK, it's two bars of E flat minor, one bar of E flat minor suspended, E flat minor seven on the next one …’ before drawing their attention to the ‘A flat suspension, and then the A flat’ (p. 205). What we know, from Marcus's book and accounts elsewhere, is that Dylan left the studio and returned the next day and recorded the song in the familiar 4:4 time and in the key of C – a far more direct pulse and the song transposed down a semitone.

However, we don't know why this happened. What made him change it? It seems such an obvious question to want to pursue, but no one seems to have asked it of Dylan or of anyone else there at the time. It is clearly easier for a guitarist to play the song in the key of C than it is in D♭/C#; the fingering on the guitar requires less effort and is far more straightforward and the chords will sound more ringing, using more open strings. But, I doubt if the reason is that simple. Al Kooper, who played organ on the session, has suggested that Bob was playing a guitar tuned to a chord of C, an unusual tuning used by a few blues musicians, and according to Kooper used first by Bukka White in the 1930s. Although admitting that he hadn't noticed in the studio at the time and deducing it later from listening to the recording, Kooper remarks: ‘it may not be apparent on the average listen, but the C tuning causes a certain frequency range to be filled that would most certainly not be if Bob was in regular tuning’ (cited in Marcus Reference Marcus2005, p. 234). However, it's very difficult to hear this and a number of accomplished guitarists that I have spoken with are sceptical. If we accept this claim we could presume that Dylan re-thought the production of the song while doodling on his guitar between the two sessions and we can speculate that it might have been the experiment with tuning that partly led him to the different arrangement. Or maybe not. At the same time, Dylan would have certainly been listening back to the tapes made in the studio – sitting alone, or with trusted confidants, analysing the recording, thinking about how the song needed to be produced, to come over on the radio, to grab people's attention as a great record. Such speculation is surely the basis for some further research into this track and session.

‘Like a Rolling Stone’ is recognised as one of Dylan's greatest songs. A study of Dylan's phonographic imagination might be developed further by considering the production of this track alongside the recording of a song that many critics consider second only to ‘Like a Rolling Stone’. Versions of ‘Blind Willie McTell’ (that song) were attempted for the album Infidels (1983). At the time Dylan was unhappy with a piano and acoustic guitar version recorded with Mark Knopfler and also with a rather insipid band arrangement. He eventually released the sparse acoustic version on the first bootleg series collection in 1991 and only began performing the song in concert during August 1997 after it had been circulated and recognised as a recording. Dylan's reluctance to initially approve the release of the studio recording is not only an indication that he was trying to achieve a specific sound, it determined that the song would not feature as part of a live performance for some years afterwards. Here the process of recording thwarted the potential of a song that was only added to Dylan's live repertoire after it had been released, received and widely recognised. No matter how he may arrange and perform his songs in concert, they will continue to carry their weight as recordings, shaped by Dylan's idiosyncratic phonographic imagination.