‘The Cruel Miller' Revisited
[The original article may be viewed at https://www.academia.edu/16917625/Memory_Print_and_Performance_The_Cruel_Miller_Revisited_
The charts at the end are not reproduced.
R. matteson 2016]
Memory, Print and Performance (‘The Cruel Miller' Revisited)
by Tom Pettitt, 2015
Cultural Sciences Institute and Centre for Medieval Literature
University of Southern Denmark
This is a draft paper delivered more informally, with focus on the main points, at the Folk Song Conference organized by the English Folk Dance and Song Society, Cecil Sharp House, London, 10 October 2015. It will in due course be replaced by a more formal version with fuller apparatus and broader contextualization. Apart from a couple of corrections, the Textual Appendix below corresponds to the session handout (and its incorporation here explains the sideways orientation of the document as a whole).
Opening Statement
This presentation will comprise a mixture of romantic generalization and intense textual analysis, which are at least designed to be mutually enlightening, to the extent indeed that the generalization may founder on the rock of the analysis.
The Song
The textual focus will be exclusively on one song (see Appendix), familiar to scholarship as Roud 263.[1] It is also known as
“The Berkshire Tragedy” / “The Wittam Miller” – the most usual titles of the original form, a long (44 quatrains) broadside ballad
first published in the first quarter of the 18th century and reprinted for up to a century thereafter (the first column of the Appendix provides one printing in full, with a collation of all the others to which I have had access, ca 12 in all) and as
“The Cruel Miller” / “The Bloody Miller” – the most usual titles of a radically revised version, a short (18 quatrains) broadside
published from the early 19th century and reprinted probably as long as broadsides continued to be published (the second column of the Appendix provides one printing in full, with a collation of all the others to which I have had access, 17 in all)
and by a variety of titles as transcribed or recorded over the last century or more from what for the moment I’ll call ‘oral tradition’ in the British Isles and North America, not to mention Tristan da Cunha (where it is known as “Maria Martini”). It figures more frequently in Scottish singing tradition than the Roud Indexes suggest, as it is mainly known in Scotland as “The Butcher Boy”, and versions with this title have inadvertently been catalogued under a quite different song with the same title, also known as Roud 409 (first line usually ‘In Jersey City ...’)[2] (the third column of the Appendix provides the full text of one version from Scottish singing tradition, while the fourth column offers a medley of possibly interesting snippets from other ‘oral’ versions from England and Scotland.[3]
The Generalizations
I have long been interested in this song partly because it is a classic ‘Murdered Sweetheart’ ballad: a girl is seduced on promise of marriage; she gets pregnant and demands the lover marry her; he instead kills her and is subsequently punished, either by supernatural intervention or more often, as here, by judicial process -- although the decisive discovery of the body in this ballad’s original form is attributed to the “watchful eye” of heaven (28.1). The issue of ballad sub-genres, their demarcation, and development is a significant aspect of folk song history, and this is one of more than thirty distinct Murdered Sweetheart ballads. Several of them are also found subsequently in ‘oral tradition’, and this one is relevant today rather because it presents a particularly awkward challenge to some of my most cherished theories about the relationships between ballads and the media by which they are transmitted over time -- in this instance through almost 300 years.
Over those centuries, prior to the emergence of the modern high technology electronic & digital media, this ballad and others have effectively been transmitted by two media systems: the medium technology of the printed broadside, and the technological zero option of the human voice and memory, aka ‘oral tradition’. And to put it crudely, there is disagreement among ballad scholars about their respective roles and relative significance. In this instance the standoff can be reformulated as: was Roud 263 a broadside ballad which from time to time was also sung, but whose printings are the mainstream of its history; or was it a song which originated in print and was from time to time reprinted but whose performances are the mainstream of its history?[4]
As it stands a common-sense approach could easily reconcile these as alternative, equally valid, perspectives, were it not that the ‘history’ of this and other songs involves not merely continuity over time but also change over time in a matter as vital as the words of which it is comprised: be it the ‘text’ of broadsides (‘text’ in what follows will mean letters than can be read), or what the Danes call the ‘word-sound’ (ordlyd) of performances?[5] The degree and kinds of change involved are registered for example in the differences between the 10 stanzas of my Scottish example and the 44 stanzas of the 18th century broadside from which it ultimately derives.[6]
Here too a common-sense approach would suggest that it’s a bit of each, and anyway does it really matter? But this verbal ‘morphing’, as John Foley called it, encompasses significant formal, aesthetic, and even generic dimensions, reflected for example in the same columns of my handout not merely in the exact locations of the gaps indicating where stanzas have been omitted (change has involved highly constructive subtraction), but also in the red letters and lines registering the generation of verbal repetitions (change has involved highly stylistic reformulation).
To me this formal, aesthetic modulation is of the essence of ballad tradition (and I sometimes wonder where Folklore as a discipline in England might be today if it had insisted on this aspect rather than pursuing social dimensions). Again to put it crudely, while broadside ballads are essential sources for the study of social conditions and popular mores and mentalities, so we can justly hail the efforts of our colleagues at the Bodleian Library and the cousins in Santa Barbara in making them available online,[7] ‘dunghills’ like ‘The Berkshire Tragedy’ are unlikely to be nominated as part of our Intangible Heritage (should the U.K. ever sign the relevant treaty), in contrast to many of the ‘traditional’ ballads, canonised by Francis James Child’s collection, which do have significant aesthetic qualities, and which do qualify as cultural achievements. And as with other songs originating in broadsides, my Scottish performance version, compared with the original, has made significant moves in that direction. In which case, assigning credit for this achievement, as between say third-rate literary hacks in the employ of urban printers within the first major form of commercial mass culture, and peasant women (mute inglorious Miltons) singing to their children by the hearth after a day’s hard work, becomes a matter of some moment.
Preliminaries to Analysis
But before engaging with the material directly it is necessary to further clarify the two sides to the confrontation.
Oral Tradition
It is becoming increasingly clear, to me at least, that ‘oral tradition’ is an inappropriate and potentially misleading term for the textless transmission of songs achieved by singers, even leaving aside the distortion introduced by the ‘oral formulaic’ school a few decades ago, which threatened to equate ‘oral’ with ‘improvised in performance’. And there problems relate to both words in the expression.
‘Tradition’ suggests a medium capable of transmitting a song over a lengthy period: if not “since time immemorial” then at least “since the later middle ages” -- in the incautious formulation by Paddy Fumerton which provoked David Atkinson’s ire in his recent book (p. 27). Atkinson’s The Anglo-Scottish Ballad and its Imaginary Contexts[8] provides powerful arguments for the primacy of broadsides in ballad transmission (and so by implication for their aesthetic upgrading) based on the undeniable ubiquity of broadsides, but he offers concessions (p. 34) to the alternative to the extent that: 1) the owner of a broadside ballad might well commit the song to memory and later perform it from memory; 2) that one singer might well learn a song from its performance by another; 3) such passing on via performance and memory might well extend through a concatenation of three such singers.
Such limitation is acceptable for present purposes, whether or not the phenomenon qualifies as ‘tradition’, not least because the actual number of singers involved is not the primary factor on the ‘oral’ side of the equation. For another awkwardness in ‘oral tradition’, this time deriving from ‘oral’, is its focus on how a song is passed on through performance from one singer to another: ‘oral’ in emerging from the mouth of the singer; ‘aural’ (the pronunciation indistinguishable in my English) in entering the ear of the listener. This process may have some role to play in verbal morphing, say if the listener mishears or misinterprets the words performed, and furthermore the notion that wordsounds that are heard are remembered differently from texts that are read is evidently current enough for David Atkinson to feel the need to question it, more than once, in his book.
But of course this is something we can never determine until we can literally read the mind of the one who has received the song, for as things stand the only way to check what they have remembered is to register a performance, which will by then however reflect the impact not merely of commitment to, but also retention in and retrieval from (recollection) the memory of the individual concerned. In our agreed chain of three singers, I contend that the verbal impact of the song’s passage through the three memories is at least significant as, and probably more significant than, the two oral-aural transfers between them: which is why I have from time to time suggested we speak rather of ‘memoral transmission’, which emphasizes this aspect, if need-be pronounced mem-óral to encompass both of them.
Broadside versus Memory
On the other side of the juxtaposition it said by some that songs on broadsides change just as much as songs in ‘oral tradition’ -- which is manifest nonsense. If you read a broadside ballad, and then read it again a week, a month or a year later, the text will be exactly the same as it was before. It is one of the definitive affordances of texts (written or printed), that no matter what the letters may get up to when you are not looking (in the case of newspapers perhaps checking out page 3), they are always back in the same places when you consult the document next time. But that is not one of the affordances of the human memory. If you instead committed the text from a broadside to memory and sang it from memory a week, a month or a year later, the words can have changed and might well have changed. As David Atkinson in another concession observes (p. 167): “although when a ballad is first learned, regardless of the nature of its source, the intention may well be to achieve a form that reproduces that source with some accuracy, subsequently that intention might become modified towards a rendering that resembles it much less closely, as the performer seeks to make the song ‘their own’”.
And should you give the broadside to someone else, when they read it, the text will be, can only be, exactly the same as when you read it, whereas if they had learnt the song from listening to you sing it and then sung it themselves, the words can have changed, and might well have changed. And ditto if the broadside is handed on to a third person, as opposed to a third person learning it from hearing it sung by the person who learned it by hearing it sung by you.
It may be objected that this is not what is meant when it is claimed that broadsides change just as much as songs transmitted through memory and performance, but it is what it should mean in comparing and contrasting the affordances of these two media. We have to go a long way to find even the possibility of change in broadside texts commensurate with the verbal morphing endemic to memoral transmission. Should two people have each acquired their own copy of a given broadside, the words of the song on these different copies would not, could not differ from each other in any way at all. It is one of the affordances specifically of the print medium (it is why printing was invented) to ensure multiple, identical copies of a given text. But this is not one of the affordances of memoral transmission, and should these two persons have instead acquired the song by hearing it sung by a third person, when they in turn came to sing it their wordsounds could differ, and might vary well have differed. The same is true, with only minor reservations, should one of these people have acquired a copy of the ballad as reissued by the same printer, or by another who has taken over his business: the text will most probably have remained ‘standing in forme’, with revision effectively a proof-reading, say correcting broken or reversed letters or manifest errors.
Textual revision in a broadside ballad is not even potentially significant until it is re-issued, effectively in a new edition by the same or more likely a different printing house, and textual changes can be introduced in relation to an existing issue used as copy-text. But the norm under such circumstances is adjustment rather than large-scale alteration. For example the undoubted changes introduced by different printers in the 12 editions consulted of “The Berkshire Tragedy”, published over roughly a century, can be quite comfortably encompassed in my transcripts which collate all the versions. With the exception of a chapbook text which omits a single stanza, all have the same number of stanzas, with the same content, in the same order, with verbal variation mainly confined within the individual stanza or indeed within the individual line. The claim that broadsides “change just as much” as memory-and-performance-borne songs is valid only for the more exceptional instances when the publication of a given song involves radical, effectively ‘literary’ revision, producing a new version, in some respects effectively a new song, as opposed to a revised edition. At which point comparison proves little: in such instances the changes in the broadside ballad could quite feasibly be greater than those achieved by the ‘systemic’ morphing inherent in memorization and retrieval from memory in performance, and should rather be compared with the interventions of singers who in von Sydow’s term are exceptionally ‘active’ tradition-bearers (like some English Gypsy singers).
But from the present – formal, aesthetic – perspective, this purely quantitative issue of the respective amounts of change characteristic of the two media systems is vastly less significant than the qualitative issue of the kinds of change characteristic of the respective media systems. Studies that have demonstrated that broadside ballads change just as much as those in memoral tradition have neglected to demonstrate that they change in the same way at least with regard the formal and aesthetic characteristics relevant here.[9]
Within my broader thesis, that verbal morphing inherent in memoral tradition can transpose sub-literary, sub-journalistic broadside ballads into the vernacular aesthetic more characteristic of Child ballads, I assert on this occasion more specifically that this is evident in the various patterns of verbal repetition discernible in the Scottish performance of this ballad and highlighted in the handout (third column), by red text and arrows -- and the illustrations in the fourth column are selected to suggest that the processes in memoral tradition generating these repetitions can either lead to alternative patterns along different transmission chains or can produce variant patternings in a song that already has them.
The Problem
This thesis has worked well enough with other ballads I’ve studied in this way, and with analogous explorations of different types of material such as spoken folk narrative (“Jack the Giant Killer”) or the ‘bad’ quartos of the plays of Marlowe and Shakespeare.[10] And if such repetition patterns are symptomatic of verbal material that has been transmitted through memory and performance, there are potentially wider implications, as they also occur in the foundational myths of religious belief systems such as the lays of the Icelandic Elder Edda and some narratives within the Old Testament.
But of course with regard to this song in particular there is an ‘elephant in the room’ which I have so far studiously avoided mentioning, and which threatens to bring down this whole house of cards. Quite simply: the nineteenth-century short broadside in the second column is manifestly a revision of the eighteenth century broadside in the first column, but it has nonetheless achieved characteristics of the kind I attribute to the impact of memoral transmission: both the creative subtraction of less than essential material and the generation of many – quite ‘traditional’ – verbal repetitions.
This has haunted me since my first published study of this ballad, from this perspective, in 1994,[11] and the current revisiting has been prompted by the discussion of this ballad by David Atkinson in his recent book,[12] precisely to document his claims for the significance of broadside revision in verbal change characteristic of ballads. It may be better than he realizes in displaying, in relation to the original broadside, nit merely change, but the kind of change I attribute to transmission through memory and performance.
Evasive Scenario
In that earlier study I offered an evasive scenario postulating the ‘traditionalisation’ of the 18th century long broadside through memoral transmission into something resembling the 19th century short broadside, the latter produced by the transcription of this song from performance rather than literate / literary revision. I was encouraged by the consideration that this scenario might be supported by the occurrence, in some few English and Scottish performances, of words or phrases which were not in the short broadside, but which evidently derived from the long broadside, in turn perhaps suggesting that it reflected, rather than accomplished, change. In the interim however their number dwindled, as I encountered short broadside variants in which those lines and phrases do occur (those that survive are highlighted in yellow on the handout). Furthermore I have the impression that the notion of broadside printers publishing songs recorded from performance is based on anecdotal, rather than documentary evidence. And the anecdote concerned is less than helpful in attributing the practice to the London printer Jeremy Catnach. For while Catnach did indeed publish a version of the short version, he also published what must have been almost the last version of the long broadside text – whose availability would have rendered recording from song tradition unnecessary. Indeed such considerations point rather to Catnach as responsible for or instigator of the revision, in the context of the highly competitive broadside trade, with both access to the original and a sense of what was needed to accommodate it to the changing forms and fashions of the broadside, including the need for reduction to fit the smaller, ‘slip’ format. Another prolific London printer of the same period, Catnach’s rival, John Pitts, also published both broadsides, so the earlier, 44 quatrain version was still very much available, but is effective eclipse in singing tradition by its new rival certainly attests the latter’s appeal.
Conciliatory Scenario
On the other hand, whoever was responsible, this was manifestly a one-off occurrence for this song. Once published, the text of “The Cruel / Bloody Miller”, through many reprintings, reverted to the relative stability characteristic of broadsides. Although change is more frequent and more extensive than in “The Berkshire Tragedy”, and can on occasion disturb the distribution of material between stanzas, the differences are in degree rather than kind, with what are effectively the same stanzas, in the same order, the collations here too capable of representation within a single column of text.
To witness again the degree of change and kinds of change achieving the revision of the long broadside into the short, it is necessary to turn to the versions recorded from singing tradition (as represented in the third and fourth columns). At the same time the evident success of the new broadside among singers suggests that its revision involved more than simple reduction to meet a new format -- including, perhaps, the generation of ‘traditional’ features analogous to those characterizing the songs they already sang and so more attractive.
Appositely, in another of his concessions David Atkinson entertains the possibility (p. 51) that broadside revision “may have been influenced by the ballad in its orally circulating form” (which would have included these features). If deliberate (as he implies), this would perhaps produce something more akin to a pastiche of a traditional ballad (say of the kind perpetrated by Sir Walter Scott), and I would prefer to envision something intermediate between conscious, literary intervention and the more systemic processes somehow inherent in transmission through memory and performance.
There may be enlightening analogies in the transmission of medieval narratives by scribal copying. Most students of medieval textuality would now concur that this was the norm for popular, late-medieval narratives in English, from Piers Plowman to the tail-rhyme romances. Nonetheless some manuscript versions of such works, which thanks to the internet are also becoming increasingly available outside the libraries where they are housed, display distinct formal symptoms, not least the verbal patternings discussed here, of transmission through memory and voice, rather than, or as well as, eye, hand and text.
In an earlier generation such signals would have prompted discussion of possible transmission by minstrels, and indeed modern
experiments in the performance of popular romances from memory have also resulted in the generation of such repetitions. But there is a viable alternative in the figure of the scribe himself, who although in most instances a literate clerk in one of the lower orders of the priesthood, was in many cases of lower class origins (these being one of the few avenues of social advance for intelligent youths), and very likely still in touch with the vernacular, memoral culture of his family and original sub-culture. Since few people can make their eyes look in two different directions simultaneously, any scribal transcription, even of the most sacred texts, had to involve committing one or more letters, words or phrases to memory from an existing text, preparatory to writing them down, from memory, to produce the new text -- and there might have been a more strictly oral (+ aural) element in the process as the scribe, as was then the norm, vocalized the words concerned. In dealing with verbal material in the vernacular, effectively in the realm of popular culture, where verbatim reproduction was non-mandatory, the original will have been consulted less frequently, the sequence of words transferred from page to page via memorizing
and recollection correspondingly longer, at some point perhaps modulating into the reverse configuration of transcription from memory, supported by occasional glances at the original, and even drifting into significant revision, these latter processes introducing features characteristic of memoral transmission.
Something similar might be imagined of a nineteenth-century printer, or a hack writer in his employ, drafting a new edition of an older broadside ballad, not least if constrained to reduce the text to fit a smaller broadside format. And perhaps it would be salutary if more generally, sub-literary revision of texts, and memoral modulation of wordsounds, were perceived not as distinct categories of ballad mediation (such categorization itself symptomatic of literacy-influenced cognition), but as theoretical extremes, somewhere between which, the realities of ballad transmission and ballad-shaping can be situated.
APPENDIX: PARALLEL TEXTS
Original Broadside: “The Berkshire Tragedy” Revised Broadside: “The Cruel Miller” Scottish Singing Tradition English / Scottish Tradition (excerpts)
[the remainder must be viewed from charts]
Footnotes:
1. Steve Roud, Roud Folk Song Indexes, http://www.vwml.org/search/search-roud-indexes?ts=1444235219442&collectionfilter=RoudFS;RoudBS&advqtext=0|rn|263#.
2 http://www.vwml.org/search/search-roud-indexes?ts=1444235609286&collectionfilter=RoudFS;RoudBS&advqtext=0|rn|409#. This other ”Butcher Boy” rejects a girl who then commits suicide, while ours, like the miller from whom he has been recast, kills a girl he has made pregnant.
3 There is an entertaining and informative review of the broadside and ’oral’ versions of the song by an author writing under the pseudonym of ’Dungbeetle’ in ”A veritable dungheap: The Bloody Miller”, English Dance and Song, 65.4 (Winter 2003): 22-23.
4 Which makes this paper, as reflected in frequent asides, a continuation on my sustained if sporadic discussion on this issue with David Atkinson, as it happens one of the organizers of this conference and indeed chairing this session. Our respective previous engagements with this particular song are touched on below.
5 The translation is mine, but I owe this felicitous suggestion to Lene Halskov Hansen’s new study of Danish ballads in performance tradition(song and dance),
Halskov Hansen, Lene. Balladesang og Kædedans. To aspekter af dansk folkevisekultur.Copenhagen: Museum Tusculanum, 2015
6 The significance of the broadsides and broadside rewriting in its Anglo-American trajectory emphasized by G.M. Laws in the essay on ”Ballad Recomposition”
forming part (pp. 105-122) of his American Balladry from British Broadsides (Philadelphia: American Folklore Society, 1957).
7. http://ballads.bodleian.ox.ac.uk/; http://ebba.english.ucsb.edu/
8. David Atkinson, The Anglo-Scottish Ballad and its Imaginary Contexts (Cambridge: Open Book Publishers, 2014)
9 Dugaw, Dianne M. “Anglo-American Folksong Reconsidered: The Interface of Oral and Written Forms”. Western Folklore. 43.2 (1984): 83-103.; Renwick, Roger de V. "The Oral Quality of a Printed Tradition". Acta Ethnographica Hungarica. 47 (2002), 81-89. It might also be objected that juxtaposing versions of an unknown relationship (e.g. which came first) demonstrates merely difference, as opposed to change and its causes: for example the differences between two arbitrarily selected
memoral versions might actually derive from their respective derivations from different broadside versions.
10 The studies concerned are available at or through https://southerndenmark.academia.edu/ThomasPettitt.
11 Pettitt, Thomas, "'Worn by the Friction of Time': Oral Tradition and the Generation of the Balladic Narrative Mode", in Contexts of Pre-Novel Narrative. The European Tradition. Ed. Roy Eriksen. Berlin & New York: Mouton de Gruyter, 1994, pp. 341-72 (an author’s draft available on academica.edu).
12 Atkinson, The Anglo-Scottish Ballad and its Imaginary Contexts, pp. 43-47.