Frequently Asked Questions: Methodology
What do you consider to be "Irish traditional" music?
Once a tune is considered to be part of the Irish repertoire by Irish traditional musicians who play that tune, then it's Irish, regardless of where it may have originated or how it may be played in other cultures. This understanding is consistent with the generally reception-oriented approach I take throughout the Tunography. Two interesting borderline examples of tunes now solidly accepted in the Irish tradition might be: the polka Farewell to Whiskey (#612), originally composed in Scotland by Niel Gow; or tunes by the Englishman James Hill such as The Scholar (#1729).
Note that I have also indexed a small number of non-Irish tunes because they are presented along with Irish traditional tunes played by Irish traditional musicians. In these cases I include a note about what culture these tunes belong to. Well-known borderline cases include Scottish tunes such as The Jig of Slurs (#936) or French-Canadian tunes such as Saint Anne's Reel (#1819). If the Irish musicians who play these tunes still think of them as "borrowed" tunes from those cultures, then I will categorize them as such here.
A tune becomes "Irish traditional" once it is played informally and privately by a body of musicians who consider themselves to be playing Irish traditional music. For a more idealistic expression of this thought, read the relevant part of the "Music" page on my personal Web site.
How is a "distinct tune" defined here?
Without wanting to get too involved in a "theory of the tune," let me make at least the following comments here. I take a fundamentally reception-oriented approach, inspired perhaps most of all by Umberto Eco's community-related comments on textual semiotics. This means that if the published evidence at hand shows that the community of Irish traditional musicians understands a tune a certain way, then that's the way I define the tune. Furthermore, I give priority to a (recorded) performance of a tune over a transcription of it. Finally, I agree with Tomás Ó Canainn's method of analysis of Irish dance tune melodies as described in his 1978 volume Traditional Music in Ireland (especially pages 27–34 – see TOC in the Bibliography). Ó Canainn's ideas have been refined and expanded by James R. Cowdery in his book The Melodic Tradition of Ireland (1990); while Cowdery indirectly challenges the possibility of an index such as this, I am nevertheless sympathetic with his demonstrations.
I have also come to my own conclusion that anyone should be reluctant to publish data on tunes which he or she does not actively play, since one cannot possibly be as confident of one's indexing work on those tunes. This is, frankly, a great weakness of all other indexing efforts to date (e.g., see links to other indexes). If you're not participating in playing the tune in the community of traditional musicians, then you're lacking much of the critical information needed to identify the tune and its relationships to the rest of the body of Irish traditional music.
Practical consequences of these theoretical positions of mine include the following rules for this index:
- You will find notes about tunes that are musically similar with direct links offered between those tune entries, to help satisfy readers who have different opinions about what a "distinct tune" is, and to avoid fruitless attempts to solve the impossible puzzles sometimes presented by the intertwined histories of similar tunes. For examples of links between similar tunes, keep reading here.
- Similar tunes which some people consider to be mere variations of a single tune will be classified as separate tunes when a particular musician has recorded both "variations" as distinct tunes. The most striking example of this rule being applied is in my distinguishing the two jigs Dan the Cobbler (#349) and Kitty's Rambles (#1059), based on the fact that Jimmy Power recorded both tunes in 1967 under these names. That seems to be a clear sign that he considered them to be distinct tunes, even though the difference lies only in a few notes. However, I don't play those tunes, so I wouldn't defend that distinction strongly. A better example of this kind of distinction is my indexing of the two related reels The Maids of Castlebar (#2379) and McFarley's Reel (#1192), which I have played with others in sessions as tunes distinct from each other, or the two versions of Boys of Tandragee (#209 and #3289) which I play as two distinct tunes. See also my note below about the Tinker tunes.
- If the above rule does not apply, then what appear to my ear to be simply variations of a tune will be listed as instances of a single tune. This allows for evolution of tunes over time. The most common example of this is in the sometimes marked differences between contemporary recordings of tunes and the older transcriptions of the same tune, such as in O'Neill's collections. I would venture the explanation that the more hands a tune has passed through, the more it evolves.
- On the other hand, if the evidence at hand indicate that a particular "variation" of a tune has begun to be played as a fixed and treasured tune, passed on from musician to musician, while the other "variation" still remains established among other traditional players, then I separate those "variations" out as musically distinct tunes. You might consider my indexing work on The Cuckoo's Nest as an example of this rule being applied (use the Tune Search and Finder). Note that I especially consider large-scale structure changes to be a sign of development of a new distinct tune, such as an entirely new part being substituted, or the entire outline of the incipit changing dramatically.
- Contrary to the above rules, but in agreement with my basic philosophy, if a book offers two closely related tunes as distinct tunes, I will still group them as instances of a single tune. O'Neill's collections offer a good example of why this is the best policy: it appears, especially in O'Neill's Music of Ireland ("1850" in this site), that O'Neill may not have realized that he already had a transcription for a given tune in some cases. Even Breathnach was susceptible to this problem – see The Old Concertina Reel (#1455). This underscores my argument that the person best qualified to publish information about tunes (including transcriptions of them!) is a musician who plays those particular tunes (see rule number 7).
- Relationships of similarity and shared history between tunes are nevertheless recorded in my database using the parenthetical notation "(compare to . . .)" at the end of the list of known titles for a tune. This allows people who disagree with my tune identity theory to see satisfaction in this index.
- Perhaps the best example of the mess that reality can make out of anyone's tune identity theory can be found in my indexing of the many variations of the closely related tunes – nearly incestuous relationships in this case! – The Yellow Tinker (#552) and The Drunken Tinker (#2088), with renditions of Dick Cosgrove's Reel fitting in as yet another tune floating between the two, but which I don't play and is therefore subsumed as variations of #552 or #2088. And never mind The Jolly Tinker (#1003), which adherents of the "tune family" theory would throw in together with all of the above tunes as just variations of a single tune. I suggest that the moral of the Tinker story is that any index author must make it clear that the identity he or she assigns to any tune may be at least subconsciously influenced by the way he or she learned the tune from the tradition and how he or she actively plays the tune within the tradition. The best way to make that clear is by providing the context of that identity in the form of a complete repertoire of the indexer at the time the tune's identity was stated. For example . . . this Web site.
Why do you transcribe just the first two bars of a tune?
I don't transcribe the entire tune. Usually I only transcribe the first couple bars or so, or at most the start of each 8-bar part. The reason I bother writing a transcription at all is to use it when I'm work on identifying tunes on recordings and in books, and to help you recognize tunes as you browse content throughout irishtune.info. The first two bars is all I need to use the software I wrote to find possible matches in my tune database (the Tunography). Then I simply sit and listen to the rest of the entire recording, or read the book, while visually or audibly comparing it to other published tune sources listed in the Tunography. Naturally this is vastly faster and easier than having to make a complete transcription.
Note that even if I personally made complete transcriptions of all tunes, I still would not have the legal right to post them without permission from known composers and arrangers. Generally even trad musicians copyright their arrangements of tunes when they publish the recordings, so noone has the legal right to freely publish a transcription of their performance. And even if the musician doesn't care, their publisher usually does.
Meanwhile, existing projects such as abc versions of O'Neill's and of other public-domain arrangements of tunes are a fine way for our community to proceed, and don't need any help from me, other than what I already do. If you need the "dots," simply scroll just a little bit further down the page on each Tune Info page to see where I've helpfully pointed you to reliable, accurate sources of published transcriptions for that tune.
Where do the tune titles come from? Who decided which title comes first?
For each tune, you see all of the titles – and only the titles – under which that tune was published in the various sources I list on that tune's Tune Info page. I present the titles exactly as they were published, even including misspellings, Irish-language titles, mislabeled tunes, and other errors that one finds in many albums and tunebooks.
- Regardless of how problematically a recording lists a tune, whether as "Untitled," or with obvious errors such as reversing the order of tunes or typographical errors, I still include these titles, since many people have no idea that these are errors and the tunes then become known among musicians by these erroneous titles.
- If a book, however, designates a tune as "gan ainm" or "untitled" I do not include that in the list of titles – see instead my index of Untitled / Gan Ainm Tunes in Books.
- When I include a note that a tune has been "mislabeled" on a particular album, that means that there was a very obvious error made such as mixing up the names of the tunes on that track. For example, if a musician records Connaughtman's Rambles and then My Darling Asleep, but the album lists these titles in reverse order, then clearly the standard titles for the tunes were intended, but an editorial error introduced this "mislabeling." In all other cases I make no assumptions about what a "standard" or "correct" title for a tune should be and present to you the published titles as they stand in these published sources.
- When I point out in my notes for a particular album that the publisher misnumbered the tracks, then I don't bother entering all the wrong titles for all those tunes. In other words, I don't treat that as a "mislabeling" error as above. So you should just look at the exact track locations I give and trust those numbers, not the numbers printed on your CD case.
- I only include titles that are not in the sources I name for you in the very rare, clear-cut cases in which I have evidence (explained in a public note with that tune) that a significant body of traditional musicians refer to the tune using that otherwise-undocumented name.
The title listed first for each tune is simply the most-often used title used in the sources listed for that tune. After that, I list titles in descending order of how many times they appear across the sources, whereby I group similar titles together in order to make the entry more readable. When tie-breaking rules are necessary, I give priority to older names over newer names, and some priority to abstract names over the personal names of musicians who are credited as sources for the tune. In a few cases historical evidence is presented in the published sources which clearly disagrees with the majority opinion, such as when the original composer disagrees with the majority, and I do let that override weak majorities. When there is still no clear priority to one title over the other, then I may finally give priority to the title I've heard being used more often for this tune in sessions. As you can see, the order is really an editorial decision rather than a codified rule. (It is not generated by a computer program).
How do I read the codes for recordings and books?
For each tune I have identified all occurrences of this tune on books and albums which I have in my possession. I don't list any other sources for the tune, because I can't verify them myself. See also: "How reliable is the data?"
Example:means "the third tune on track seven."
The first number is the track number. The second number tells you whether this tune is played as the first, second, or third, etc. distinct tune on that track. For some longer tracks, such as Kevin Burke's infamous track 8 of If the Cap Fits, I've also given the exact time at which the particular tune begins. Background note for those unfamiliar with folk music: Usually Irish dance tunes are played in a sequence of different tunes connected seamlessly, without dropping a beat, in what we call a "set of tunes."
This track#tune location information is especially valuable in the many cases where the published album contents information is incomplete, misnumbered, or totally wrong! The tune locations I publish here are more accurate than the original printed matter published with the album. The track numbers I give you match the actual numbers on your CD player.
I also assign an "album code" to every album in the index. You can get an index of the album codes in the Discography.
For every transcription I have found which matches this tune, you see a book code (which is explained in the Bibliography) plus a number indicating where to find this tune in that book. The "#" symbol precedes tune numbers, if that book gave tune numbers. If you don't see a "#" symbol, then the number I give is a page number. For example: "Cr 152" means page 152 of Matt Cranitch's fiddle tutor, whereas "1850 #1178" means tune number 1178 in O'Neill's Music of Ireland.
Each tune's "Recordings" and "Books" sources are displayed in chronological order. The order of the Recordings codes does consider the particular original recording dates of individual tunes. Similarly, in books where each tune is provided with an original source date (such as CRE 2), that date is considered in the sort order here. The idea is to send you to the oldest sources first, in accordance with my principle that "older is better" when more personalized criteria don't apply – see Tip #6 in Tips for Learning Irish Traditional Music.
How reliable is the data?
The basic structure of the database guarantees that:
- Each recorded or notated setting of a tune is linked to only one distinct tune. I.e., no tune setting, whether on a recorded track or in a book with numbered tunes, can be cited more than once within the entire database.
I have also consistently held myself to the following rules:
- The identity of every tune recording, and of every tune transcription in books with numbered tunes (e.g. "1001 #398"), has been verified (see next item) by me. I also guarantee my identifications of transcriptions in books which fail to include tune numbers (such as Cr, FF, or M2) if I provide at least a page number location.
- "Personally verified" means that I never trust anyone else's information about a tune, whether album liner notes, scholarly publications, or personal communications – although I may use those sources as leads for me to then verify. It means that I have listened to the recording in question, or read the transcription in question, and compared it simultaneously with the other recordings and transcriptions to which I have linked it. I do not rely on any sort of automation or "formula" (although I may use them to give me preliminary guesses). Currently there are over 250 million potential relationships between all of the tune recordings and transcriptions in my database, and if I enter the exact location as above, then I am willing to defend my indexing work as if I had carefully considered all those millions of possibilities and arrived at the best solution. However, as discussed here, my definition of "same tune" is slightly more relaxed regarding tunes I don't play. If you would like to debate one of my verified links, first follow the exact references I give and judge for yourself! I'd be very happy if you can prove me wrong, so that I could correct my error. See who else has done exactly that!
- See also the rules that I describe for each field:
What is the tune ID number?
The number in the column labeled "ID#" is a number I have permanently assigned to each distinct tune. I may abandon use of a particular ID number if I find a better way to index that tune, but I will never reuse that number for a different tune. I provide this number throughout the site to allow you to cite accurately information about particular tunes from this index in your on-line discussions or in your papers and publications (see How should I cite this Web site?).