Textual Variation and Selection in Post-Medieval Icelandic ulur

Slide Note

This presentation discusses the challenges in structuring post-medieval Icelandic ulur, a genre of folk poetry, and proposes a possible solution using algorithms. It explores the variability and instability of ulur texts, highlighting the difficulty in defining them as poems. The speaker, Yelena Sesselja Helgadóttir, presents her methodology, preliminary results, and conclusions on representative text selection. The project focuses on the composition, performance, and metrical form of ulur, emphasizing the intertwining nature of structure, composition, and performance in Icelandic folklore.

christel Follow

Uploaded on Sep 16, 2024 | 0 Views

Download Presentation

Please find below an Image/Link to download the presentation.

The content on the website is provided AS IS for your information and personal use only. It may not be sold, licensed, or shared on other websites without obtaining consent from the author. Download presentation by click this link. If you encounter any issues during the download, it is possible that the publisher has removed the file from their server.

E N D

Presentation Transcript

Textual Variation and Representative Selection of Texts The Case of Post-Medieval Icelandic ulur Yelena Sesselja Helgad ttir (Stofnun rna Magn ssonar slenskum fr um The rni Magn sson Institute for Icelandic Studies) Plotting Poetry 5 Popular Voices Estonian Literary Museum, Tartu 4-6 July 2022

This presentation Introduction postmedieval Icelandic ulur my current project (in general) Methodology challenges the structure of postmedieval Icelandic ulur and text selection (possible solution) suggested algorithm and questions to it discussion Preliminary results & conclusions Tartu, 6 July 2022: Plotting Poetry 5 Yelena Sesselja Helgad ttir: Textual Variation and Representative Selection of Texts 2

Postmedieval Icelandic ulur (PM) Postmedieval Icelandic ulur (not to be confused with Old Norse ulur!) a genre of folk poetry, ca. 15th 20th c. versified, but not stanzaic lists of names, sequences of short motifs and/or longer narrative episodes in very free poetic form relatively short (25 lines at average), fragmentary and intersecting highly variative! due to their instability even at the very basic structural level which makes their designation as poems problematic, since in many cases there is no text (or group of texts) that is stable enough to be labelled a poem text A contains stable structural units abcd, text B contains units abdef; can they be identified as the same poem? this makes PM more variative than the majority of Icelandic folk poetry The majority of manuscript texts of PM have neither be edited nor digitized; audio archives are available at Ismus.is Tartu, 6 July 2022: Plotting Poetry 5 Yelena Sesselja Helgad ttir: Textual Variation and Representative Selection of Texts 3

Structure, composition, performance In this project, I work with the structure & composition of ulur research of their metrical form, also using quantitative methods, is in my slenskar ulur s ari alda. PhD diss., School of Humanities, University of Iceland. Reykjav k. 406 pp. (In Icelandic; https://hdl.handle.net/20.500.11815/1939) Formulaic Language in Minimal Metrical Requirements , in: The Ecology of Metre: A special issue of RMN Newsletter (11). Ed. Ilya Sverdlov & Frog. Helsinki. Pp. 49 61. Structure, composition and performance are intertwined as other folklore, ulur are composed in each performance performers of ulur make extensive use of the structural units of ulur in their composition in performance as shown in detail in my PhD diss., esp. chapter 8.4-8.5, pp. 240-249 Tartu, 6 July 2022: Plotting Poetry 5 Yelena Sesselja Helgad ttir: Textual Variation and Representative Selection of Texts 4

My current project Textual Variation and Representative Selection of Texts: The Case of Post-Medieval Icelandic ulur supported by RANN S Icelandic Research Fund, grant 228363-051 individual project: I am the project leader and its only staff member I am closer to the beginning all results are preliminary! It aims at finding a scientific solution to the methodological question of effective text choice for the purpose of research or scholarly editing for folk poetry where texts are relatively short, but highly variable and many poems have no clear boundaries the main challenge is to find an algorithm for producing a manageable, but representative selection that reflects the specificity of the genre in question Tartu, 6 July 2022: Plotting Poetry 5 Yelena Sesselja Helgad ttir: Textual Variation and Representative Selection of Texts 5

Specificity of PM ulur Specificity of PM ulur lies in their loose and highly variable two-level structure based on the listing principle the lists at the first (lower) level consist mainly of two kinds of building units heiti: names and other nouns repeatedly appearing in ulur motifs: simple, repeated descriptions of various kinds of events, actions, conditions, etc. these lists, combined with single names and/or motifs, produce the main second-level unit: blocks : relatively stable combinations of heiti (names) and/or motifs and, eventually, minor blocks some blocks combine into larger and less stable units designated as block sequences PM -texts can be made of units from both levels Text: blocks heiti, motifs Tartu, 6 July 2022: Plotting Poetry 5 Yelena Sesselja Helgad ttir: Textual Variation and Representative Selection of Texts 6

Structure of PM ulur an illustration Minni = motif Yfirblokk = block sequence Tartu, 6 July 2022: Plotting Poetry 5 Yelena Sesselja Helgad ttir: Textual Variation and Representative Selection of Texts 7

An example of PM (B 816 8vo 109r, Hvarfsbk) Bukonann dillade b rnunum b dum J runne runne ngunne Kingunne Adalvarde ormagarde, Eijrijke og Sveine, mu<n>du fr n<d>urnar ijnar Kambur Sk[i] rings son, Ski ringur Brands son Brandur bi rg lfs Son Bi[ ]rgolfur Hrings Son [...] Hake denns son allra Jllra Tr lla fader i Hellre The housewife lulled both [sic] children J runn<,> runn<,> Ingunn<,> [Kingunn]<,> A alvar ur<,> [Ormagar ur], Eir kur and Sveinn, remember your relatives [Kambur]<,> son of Sk ringur, Sk ringur<,> son of Brandur<,> Brandur<,> son of Bj rg lfur<,> Bj rg lfur<,> son of Hringur [ ] Haki<,> son of inn<,> the father of all trolls in a cave introductory motif block block name list block sequence introductory motif block name list Tartu, 6 July 2022: Plotting Poetry 5 Yelena Sesselja Helgad ttir: Textual Variation and Representative Selection of Texts 8

An example of PM (Holm papp 64 fol 99v) g sat undir fiskahla a f ur m ns, menn komu a m r, r ku staf hnakka m r, hla inn t k a brenna, g f kk a renna upp biskups land, biskup tti gott b , gaf m r uxa og k , Uxinn t k a vaxa, k rinn t k a mj lka, Sankti Mar a gaf m r sau , a var m r a miklum au , etc. I sat under my father s fish stack, men came to me, stroke me with a stick onto my neck, the stack started burning, I got (/was allowed) to run to <the> bishop s land, <the> bishop had a good farm, <he> gave me a bull and a cow, the bull started growing, the cow started milking, <the> Virgin Mary gave me a ram, it turned out to be big wealth for me, etc. block = list of motifs (each motif usually corresponds to 1-2 lines) Tartu, 6 July 2022: Plotting Poetry 5 Yelena Sesselja Helgad ttir: Textual Variation and Representative Selection of Texts 9

The listing structure and text selection This listing structure is one of the most specific features of PM and the only of the specific features of PM that would be underrepresented in selections of texts governed by chronological, geographical, topical or other traditional principle(s) This can be remedied by making the structure central in selecting texts In the case of folk poetry with unclear poem boundaries, but chiefly consisting of a limited number of structural units (such as PM -blocks) which have clearer boundaries than the poems, a manageable and representative selection can be based on the typology of these units rather than on a somewhat fuzzy typology of the poems The solution is to make the PM structure into a tool for selecting texts Tartu, 6 July 2022: Plotting Poetry 5 Yelena Sesselja Helgad ttir: Textual Variation and Representative Selection of Texts 10

Text selection criteria (my initial argument) The text selection which most adequately represents a useful overview of highly variable texts such as PM , while also reflecting their specificity, should be based on the most typically encountered variant(s) of each structural unit of PM (primarily, of each block) temporal and geographical considerations should then be applied as second and third criteria of text selection in the case of PM , the very minimal-size representative selection can be based on one text of each of the roughly 50 blocks and their subtypes already identified in my dissertation; each text should contain the most common variant of the respective block the most common (most typically encountered) variants are identified using frequency analysis Tartu, 6 July 2022: Plotting Poetry 5 Yelena Sesselja Helgad ttir: Textual Variation and Representative Selection of Texts 11

35 15 10 12 12 14 12 abcd ehijklmoprstvwy abcd ehijklmoprstvwy abcd ehijklmoprstvw abcd ehijklmoprstvxy abcd ijklmoqrstvxy ab hijklmoprstvx abeijkln prstvx Frequency analysis 1. Encoding the block in question throughout the PM text corpus I use (modified) methods of manually preparing texts in Excel for phylo- genetic software analysis (I do not for a number of reasons use the software itself at this stage) each sub-unit (motif, name, sometimes sub-block) or its variant is allocated a letter the block to the right, e.g., is described as I 4: abehkl prstvx (I 4 = unique text number) 2. Sorting and counting (also in Excel) g sat undir fiskahla a f ur m ns, menn komu a m r, r ku staf hnakka m r, hla inn t k a brenna, g f kk a renna upp biskups land, biskup tti gott b , gaf m r uxa og k , Uxinn t k a vaxa, k rinn t k a mj lka, Sankti Mar a gaf m r sau , a var m r a miklum au , etc. Sat g undir fiskahla a f ur m ns. f ur m ns og m ur tti g a gj ta b s og barna, sv ns og sau a. Menn komu a m r ar kom Madur ad m r, a b c d e Tartu, 6 July 2022: Plotting Poetry 5 Yelena Sesselja Helgad ttir: Textual Variation and Representative Selection of Texts 12

Identifying variants Sankti Maria gaf m r sau , v w s an l h n steindau x hann var m r a miklum au Sub-unit variants: at this stage, at the higher level of PM structure, only major variants of motifs are considered minor variants (within motifs) will be considered at the next stage of this project: similar analysis of building units at the lower structural level of particular importance are variants that correlate with different ways to continue the text Example: vw in the block Sat g undir fiski- hla a correlates with Kona m n kofanum My wife in the hut as next block in the block sequence vx in Sat g undir fiskihla a correlates with Sofa, sofa hj nakornin b i <They> sleep, sleep both husband and wife v = Saint Mary gave me <a> ram w = then she lay stone-dead [sic] x = it became great wealth for me Tartu, 6 July 2022: Plotting Poetry 5 Yelena Sesselja Helgad ttir: Textual Variation and Representative Selection of Texts 13

vx, vw and subsequent blocks SAT G UNDIR FISKIHLA A I sat under my father s fish vx: Sankti Mar a gaf m r sau , hann var m r a miklum au Saint Mary gave me <a> ram, it became great wealth for me SOFA, SOFA HJ NAKORNIN B I <They> sleep, sleep both husband and wife stack vw: Sankti Mar a gaf m r sau , s an l h n steindau Saint Mary gave me <a> ram, then she lay stone-dead [sic] KONA M N KOFANUM My wife in the hut (Paradoxically, virtually nothing in the meaning of the subsequent blocks correlates with the Virgin Mary giving a ram (v) and then falling dead (w), or with the ram becoming great wealth (x). This correlation with the subsequent blocks seems to be exclusively a matter of structure.) Tartu, 6 July 2022: Plotting Poetry 5 Yelena Sesselja Helgad ttir: Textual Variation and Representative Selection of Texts 14

Step 1 End of step 1 Steps 2-4 Input: List of 50+ blocks Algorithm proposed no yes Similar procedures for identifying the most frequent variants for other building units of PM : block sequences, essential motifs and combinations of names (heiti). Output: Corpus Bm , Corpus Bh , and Corpus Bs , respectively Block 1 Other blocks? One Input: PM corpus subtype? no no yes yes Other subtypes? Identify the most frequent subtype (frequency analysis) Output: Output: Corpus A* Corpus A* Other texts: Move to Corpus A* Output: Corpus A Corpus A Output: Step 5 Block 1:1 (1:2 ) The oldest text: move to Corpus A Output of Steps 2-4 compared with Corpus A. Repetitively occurring texts detected. Corpus A updated with non-repetitive texts from Bm, Bh and Bs in order to maximize the number of motifs, name lists and block sequences represented in the corpus that is sought for. Identify all PM texts containing Block 1(:1) Identify the oldest PM text containing the most frequent variant of Block 1(:1) Output: The corpus of PM texts containing Block 1(:1(2, 3 )) Output: The corpus of PM texts containing the most frequent variant of Block 1(:1) Identify all PM texts containing that variant of Block 1(:1(2, 3 )) Identify the most frequent variant of Block 1(:1) (fr.anal.) Tartu, 6 July 2022: Plotting Poetry 5 Yelena Sesselja Helgad ttir: Textual Variation and Representative Selection of Texts 15

Some questions The main question is apparently to which extent the corpora from steps 2-4 (Bm, Bh, and Bs) will overlap with Corpus A. I assumed that the corpus, resulting from the analysis of all the blocks, would also be representative of most PM block sequences (larger structural units), and some motifs and frequently occurring names combinations (lower lever units) One and the same PM -text, composed of different blocks, might be considered as being representative of more than one block how often this will occur is another question this might give possibilities to decrease the minimal number of texts required to represent all blocks Tartu, 6 July 2022: Plotting Poetry 5 Yelena Sesselja Helgad ttir: Textual Variation and Representative Selection of Texts 19

Selecting most representative texts of a block is usually easy, although even the block Sat g undir fiskihla a, that has relatively little variation, has four different subtypes some of them correlate with a certain subsequent block other with certain period, etc. The block Sat g undir fiskihla a, however, is only the first block in the eponymous block sequence abcd {e/ }(h)ijklmo{p/q}rstvxy ( tti g a gj ta + hann var m r a miklum au +) abeijkln prstvx (hann var m r a miklum au ) SAT G UNDIR FISKIHLA A abcd ehijklmoprstvwy ( tti g a gj ta + s an l h n steindau ) ab /ehijkl{mop}/{n p}rstvx (rak/r ku staf hnakka m r + hann var m r a miklum au ) Tartu, 6 July 2022: Plotting Poetry 5 Yelena Sesselja Helgad ttir: Textual Variation and Representative Selection of Texts 20

The block sequence Sat g undir fiskihlaa abcd {e/ }(h)ijklmo{p/q}rstvxy ( tti g a gj ta + hann var m r a miklum au +) F r g til ... / Gott tti m r t a l ta SOFA HLJ NAKORNIN B I abeijkln prstvx (hann var m r a miklum au ) LEIT G UPP TIL HIMNA KOM G AR A KVELDI g tti m r ... hest/hund SAT G UNDIR FISKIHLA A (g)h: F r g til sj var / Hva ger i go i abcd ehijklmoprstvwy ( tti g a gj ta + s an l h n steindau ) (a)b: Leit g upp til himna / Sat g undir stofuvegg (f)K: g tti m r ... hest/hund + Kom g ar a kveldi ab /ehijkl{mop}/{n p}rstvx (rak/r ku staf hnakka m r + hann var m r a miklum au ) KONA M N KOFANUM / KONA B NDA GEKK TIL BRUNNS [endir] HVA ER R, BOLI MINN S KTU ti stendur t kin t ni VATNI VIGGA Tartu, 6 July 2022: Plotting Poetry 5 Yelena Sesselja Helgad ttir: Textual Variation and Representative Selection of Texts 21

Text selection for the block sequence To obtain representative texts for each subtype of the block sequence, it is desirable to find a representative text for each of the following: the subtype structure: most typical combination of blocks & connecting motifs each of the blocks (and motifs) in the subtype F r g til ... / Gott tti m r t a l ta SOFA HLJ NAKORNIN B I SAT G UNDIR FISKIHLA A g tti m r ... hest/hund LEIT G UPP TIL HIMNA KOM G AR A KVELDI SAT G UNDIR FISKIHLA A [endir] SAT G UNDIR FISKIHLA A KONA M N KOFANUM / KONA B NDA GEKK TIL BRUNNS (The other subtype of this third line occurs quite seldom and is not considered here.) (As you see, at this point I don t even think of blocks and motifs subtypes.) Tartu, 6 July 2022: Plotting Poetry 5 Yelena Sesselja Helgad ttir: Textual Variation and Representative Selection of Texts 22

D075 D238 6570 6565 I 67 II 373 I 86 I 47 s 20116 s 19473 s 19785 s 19685 D073 s 20385 s 19453 s 2572 s 20444 D225 D150 II 249 abeijkln prstvx (hann var m r a abcd {e/ }(h)ijklmo{p/q}rstvxy ( tti g a gj ta I 53 II 378 (II 28) II 249 s 19785 D075 II 284 s 19421 s 19824 D150 022 s 20116 7187-I 6570 6565 s 19453 D238 6919 (II 268) s 19880 6686 s 20444 s 19685 I 39 I 94 II 75 II 137 C002 C082a 016 (II 203)(C038) C100 C187 II 225 C038 045 D231 I 39 C002 II 178 II 249 I 47 D231 022 s 20385 s 19880 s 19785 s 19824 I 67 C182 I 94 C038 II 59 C100 II 268 046 II 75 D075 D226 II 137 F r g til ... / Gott tti m r t a l ta SOFA HLJ NAKORNIN B I I 125 060 C187 II 53[d]II 104 II 172 II 407 C004 LEIT G UPP TIL HIMNA KOM G AR A KVELDI SAT G UNDIR FISKIHLA A g tti m r ... hest/hund D030 ab /ehijkl{mop}/{n p}rstvx (rak/r ku staf hnakka m r + hann var m r a miklum au ) abcd ehijklmoprstvwy ( tti g a gj ta + s an l h n steindau ) II 394 D030 s 19888 6565 C095 s 20124 6570 I 125 II 271 D070 s 20452 I 75 II 378 D266 s 19891 I 118 C004 II 394 D262 II 225 I 86 (I2-D 83) II 271 s 19745 [og etv. fyrrv. -46?] II 238 II 31 II 68 s 19891 II 407 II 172 051 C004 C120 (g)h: F r g til sj var / Hva s 21699 D008 6807 6427 s 20566 D052 6893 6481 s 257[b] D114 6608 6463 s 2245D123 7137 7501 s 2277D126 7794 6547 I 64 D163 8411 6842 s 80 D190 7096 6600 s 10935 D201 6905 6451 II 293 D302 ( 6708) D283 II 238 046 II 407 II 172 I 86 C004 C082b I 67 060 I 125 II 104 D218 [sbr. D089] D225 6542 D262 6848 C033 II 186 II 39 016 6593 D303 6630 6848 (a)b: Leit g upp til himna / Sat g Here is our block sequence and representative texts for each block, respectively (Corpus A*); the numbers are text IDs. D200 II 190 D030 s 19550/19551 II 247 D078 s 19793 s 20124 D020 D305 s 19888 6570 (f)K: g tti m r ... hest/hund + Kom KONA M N KOFANUM / KONA B NDA GEKK TIL s 2599 s 2245 s 20420 s 172 s 80 s 10935 D008 D052 D126 D283 D302 D234 D274 D165 D312 [endir] s 2245 D008 6469 6462 6547 6430 8184 6708 D234 D052 9330 6481 6567 6822 6851 [ 6600] s 80 D165 s 21652 D145 II 284 7501 II 293 I 64 D123 s 257 D163 s 226 II 334 S KTU VATNI VIGGA HVA ER R, BOLI ti stendur t kin t ni Tartu, 6 July 2022: Plotting Poetry 5 Yelena Sesselja Helgad ttir: Textual Variation and Representative Selection of Texts 23

D075 D238 6570 6565 I 67 II 373 I 86 I 47 s 20116 s 19473 s 19785 s 19685 D073 s 20385 s 19453 s 2572 s 20444 D225 D150 II 249 abcd {e/ }(h)ijklmo{p/q}rstvxy ( tti g a gj ta + hann var m r a miklum au +) abeijkln prstvx (hann var m r a miklum au ) I 53 II 378 + abcd {e/ }(h)ijklmo{p/q}rstvxy (II 28) II 249 s 19785 D075 II 284 s 19421 s 19824 D150 022 s 20116 7187-I 6570 6565 s 19453 D238 6919 (II 268) s 19880 6686 s 20444 s 19685 I 39 I 94 II 75 II 137 C002 C082a 016 (II 203)(C038) C100 C187 II 225 C038 045 D231 I 39 C002 II 178 II 249 I 47 D231 022 s 20385 s 19880 s 19785 s 19824 I 67 C182 I 94 C038 II 59 C100 II 268 046 II 75 D075 D226 II 137 F r g til ... / Gott tti m r t a l ta SOFA HLJ NAKORNIN B I I 125 060 C187 II 53[d]II 104 II 172 II 407 C004 LEIT G UPP TIL HIMNA KOM G AR A KVELDI g tti m r ... hest/hund SAT G UNDIR FISKIHLA A D030 ab /ehijkl{mop}/{n p}rstvx (rak/r ku staf hnakka m r + hann var m r a miklum au ) abcd ehijklmoprstvwy ( tti g a gj ta + s an l h n steindau ) II 394 D030 s 19888 6565 C095 s 20124 6570 I 125 II 271 D070 s 20452 I 75 II 378 D266 s 19891 I 118 C004 abcd ehijklmoprstvwy s 21699 D008 6807 6427 s 20566 D052 6893 6481 s 257[b] D114 6608 6463 s 2245D123 7137 7501 s 2277D126 7794 6547 I 64 D163 8411 6842 s 80 D190 7096 6600 s 10935 D201 6905 6451 II 293 D302 ( 6708) D283 Here is our block sequence and representative texts for each block, respectively (Corpus A*); texts consti- tuting Corpus A are framed. (The numbers are text IDs.) + ab /ehijkl{mop}/{n p}rstvx II 238 046 II 407 II 172 I 86 C033 C004 II 186 C082b I 67 060 I 125 D200 II 104 D218 [sbr. D089] II 394 D262 II 225 I 86 (I2-D 83) II 271 s 19745 [og etv. fyrrv. -46?] II 238 II 31 II 68 s 19891 II 407 II 172 051 C004 C120 (g)h: F r g til sj var / Hva ger i go i D225 6542 D262 6848 The problem is that we can not always trace the same oldest text throughout the line: we can with D075, but not with I 39 or many other cases. Moreover, the question is whether we can always find the same text throughout the line? II 39 016 6593 D303 6630 6848 (a)b: Leit g upp til himna / Sat g undir stofuvegg II 190 D030 s 19550/19551 II 247 D078 s 19793 s 20124 D020 D305 s 19888 6570 (f)K: g tti m r ... hest/hund + Kom KONA M N KOFANUM / KONA B NDA GEKK TIL s 2599 s 2245 s 20420 s 172 s 80 s 10935 D008 D052 D126 D283 D302 D234 D274 D165 D312 [endir] s 2245 D008 6469 6462 6547 6430 8184 6708 D234 D052 9330 6481 6567 6822 6851 [ 6600] s 80 D165 s 21652 D145 II 284 7501 II 293 I 64 D123 s 257 D163 s 226 II 334 S KTU VATNI VIGGA HVA ER R, BOLI ti stendur t kin Tartu, 6 July 2022: Plotting Poetry 5 Yelena Sesselja Helgad ttir: Textual Variation and Representative Selection of Texts 24

D075 D238 6570 6565 I 67 II 249 abeijkln prstvx (hann var m r a miklum au ) abcd {e/ }(h)ijklmo{p/q}rstvxy ( tti g a gj ta + hann var m r a miklum au +) II 373 I 86 I 47 I 53 II 378 s 20116 s 19473 s 19785 s 19685 (II 28) II 249 s 19785 D075 D073 s 20385 s 19453 s 2572 I 39 C100 II 284 s 19421 s 19824 D150 s 20444 D231 D150 C038 045 I 94 C187 022 s 20116 7187-I 6570 6565 s 19453 D238 D231 II 75 II 225 I 39 I 67 I 94 II 59 II 75 II 137 F r g til ... / Gott tti m r t a l ta II 137 C002 C082a 016 (II 203)(C038) C002 C182 C038 C100 D075 D226 SOFA HLJ NAKORNIN B I 6919 (II 268) s 19880 6686 II 178 II 249 I 47 II 268 046 s 20444 s 19685 D231 022 s 20385 s 19880 s 19785 s 19824 I 125 II 53[d]II 104 II 172 II 407 LEIT G UPP TIL HIMNA KOM G AR A KVELDI g tti m r ... hest/hund SAT G UNDIR FISKIHLA A 060 C187 C004 D030 ab /ehijkl{mop}/{n p}rstvx (rak/r ku staf hnakka m r + hann var m r a miklum au ) abcd ehijklmoprstvwy ( tti g a gj ta + s an l h n steindau ) II 394 D030 s 19888 6565 C095 s 20124 6570 II 394 D262 II 225 I 86 (I2-D 83) C004 (g)h: F r g til sj var / Hva ger i go i s 21699 D008 6807 6427 II 238 046 II 271 s 19745 [og etv. fyrrv. -46?] II 238 II 31 C120 I 125 II 271 D070 s 20452 s 20566 D052 6893 6481 s 257[b] D114 6608 6463 s 2245D123 7137 7501 s 2277D126 7794 6547 I 64 D163 8411 6842 D190 7096 6600 s 10935 D201 6905 6451 II 293 D302 ( 6708) D283 II 407 II 172 II 68 s 19891 II 407 II 172 051 I 75 II 378 D266 s 19891 I 86 C033 I 118 C004 D225 6542 C004 II 186 II 39 6848 D262 6848 (a)b: Leit g upp til himna / Sat g undir stofuvegg C082b I 67 060 016 6593 D303 6630 All connections of the respective representative texts showing, among other things, the correlation of vx, vw and the following blocks. D200 s 80 I 125 II 104 D218 [sbr. D089] II 190 D030 s 19550/19551 II 247 D078 s 19793 s 20124 D020 D305 s 19888 6570 (f)K: g tti m r ... hest/hund + Kom g ar a kveldi KONA M N KOFANUM / KONA B NDA GEKK TIL s 2599 s 2245 s 20420 s 172 [endir] s 80 s 10935 D008 D052 D126 D283 D302 D234 s 2245 D234 s 80 II 334 D274 D165 D312 D008 D052 D165 s 21652 6469 9330 D145 II 284 S KTU VATNI VIGGA HVA ER R, BOLI MINN ti stendur t kin t ni 6462 6547 6430 8184 6708 6481 6567 6822 6851 [ 6600] 7501 II 293 I 64 s 257 D163 s 226 D123 Tartu, 6 July 2022: Plotting Poetry 5 Yelena Sesselja Helgad ttir: Textual Variation and Representative Selection of Texts 25

D075 D238 6570 6565 I 67 II 249 abeijkln prstvx (hann var m r a miklum au ) abcd {e/ }(h)ijklmo{p/q}rstvxy ( tti g a gj ta + hann var m r a miklum au +) II 373 I 86 I 47 I 53 II 378 s 20116 s 19473 s 19785 s 19685 (II 28) II 249 s 19785 D075 D073 s 20385 s 19453 s 2572 I 39 C100 II 284 s 19421 s 19824 D150 s 20444 D231 D150 C038 045 I 94 C187 022 s 20116 7187-I 6570 6565 s 19453 D238 D231 II 75 II 225 I 39 I 67 I 94 II 59 II 75 II 137 F r g til ... / Gott tti m r t a l ta II 137 C002 C082a 016 (II 203)(C038) C002 C182 C038 C100 D075 D226 SOFA HLJ NAKORNIN B I 6919 (II 268) s 19880 6686 II 178 II 249 I 47 II 268 046 s 20444 s 19685 D231 022 s 20385 s 19880 s 19785 s 19824 I 125 II 53[d]II 104 II 172 II 407 LEIT G UPP TIL HIMNA KOM G AR A KVELDI g tti m r ... hest/hund SAT G UNDIR FISKIHLA A 060 C187 C004 D030 ab /ehijkl{mop}/{n p}rstvx (rak/r ku staf hnakkam r abcd ehijklmoprstvwy ( tti g a gj ta II 394 D030 s 19888 6565 C095 s 20124 6570 I 125 II 271 D070 s 20452 + ab /ehijkl{mop}/{n p}rstvx + hann var m r a miklum au ) abcd ehijklmoprstvwy + s an l h n steindau ) II 394 D262 II 225 I 86 (I2-D 83) C004 (g)h: F r g til sj var / Hva ger i go i s 21699 D008 6807 6427 II 238 046 II 271 s 19745 [og etv. fyrrv. -46?] II 238 II 31 C120 s 20566 D052 6893 6481 s 257[b] D114 6608 6463 s 2245D123 7137 7501 s 2277D126 7794 6547 D163 8411 6842 D190 7096 6600 s 10935 D201 6905 6451 II 293 D302 ( 6708) D283 If we only look at the cases where there is a represen- tative text both for the sub- type structure and for each of the blocks in the subtype I 75 II 378 D266 s 19891 II 407 II 172 II 68 s 19891 II 407 II 172 051 I 118 C004 D225 6542 D262 6848 I 86 C033 C004 II 186 II 39 6848 (a)b: Leit g upp til himna / Sat g undir stofuvegg C082b I 67 016 6593 D303 6630 we can find them in most cases (blue lines), even in difficult ones (purple line) but clearly not for all block subtypes (orange lines). I 64 060 D200 s 80 I 125 II 104 D218 [sbr. D089] II 190 D030 s 19550/19551 (f)K: g tti m r ... hest/hund + Kom g ar a kveldi II 247 D078 s 19793 s 20124 D020 D305 s 19888 6570 KONA M N KOFANUM / KONA B NDA GEKK TIL s 2599 s 2245 s 20420 s 172 [endir] s 80 s 10935 D008 D052 D126 D283 D302 D234 s 2245 D234 s 80 II 334 D274 D165 D312 D008 D052 D165 s 21652 HVA ER R, BOLI MINN S KTU VATNI VIGGA 6469 6462 6547 6430 8184 6708 9330 6481 6567 6822 6851 [ 6600] D145 II 284 7501 II 293 I 64 s 257 D163 s 226 ti stendur t kin D123 Tartu, 6 July 2022: Plotting Poetry 5 Yelena Sesselja Helgad ttir: Textual Variation and Representative Selection of Texts 26

D075 D238 6570 6565 I 67 II 249 abeijkln prstvx (hann var m r a miklum au ) abcd {e/ }(h)ijklmo{p/q}rstvxy ( tti g a gj ta II 373 I 86 I 47 I 53 II 378 + abcd {e/ }(h)ijklmo{p/q}rstvxy + hann var m r a miklum au +) s 20116 s 19473 s 19785 s 19685 (II 28) II 249 s 19785 D075 D073 s 20385 s 19453 s 2572 I 39 C100 II 284 s 19421 s 19824 D150 s 20444 D225 D150 C038 045 I 94 C187 022 s 20116 7187-I 6570 6565 s 19453 D238 D231 II 75 II 225 I 39 I 67 I 94 II 59 II 75 II 137 F r g til ... / Gott tti m r t a l ta II 137 C002 C082a 016 (II 203)(C038) C002 C182 C038 C100 D075 D226 SOFA HLJ NAKORNIN B I 6919 (II 268) s 19880 6686 II 178 II 249 I 47 II 268 046 s 20444 s 19685 D231 022 s 20385 s 19880 s 19785 s 19824 I 125 II 53[d]II 104 II 172 II 407 LEIT G UPP TIL HIMNA KOM G AR A KVELDI g tti m r ... hest/hund SAT G UNDIR FISKIHLA A 060 C187 C004 D030 ab /ehijkl{mop}/{n p}rstvx (rak/r ku staf hnakka m r abcd ehijklmoprstvwy ( tti g a gj ta II 394 D030 s 19888 6565 C095 s 20124 6570 I 125 II 271 D070 s 20452 + ab /ehijkl{mop}/{n p}rstvx + hann var m r a miklum au ) abcd ehijklmoprstvwy + s an l h n steindau ) II 394 D262 II 225 I 86 (I2-D 83) C004 (g)h: F r g til sj var / Hva ger i go i s 21699 D008 6807 6427 II 238 046 II 271 s 19745 [og etv. fyrrv. -46?] II 238 II 31 C120 s 20566 D052 6893 6481 s 257[b] D114 6608 6463 s 2245D123 7137 7501 s 2277D126 7794 6547 I 64 D163 8411 6842 D190 7096 6600 s 10935 D201 6905 6451 II 293 D302 ( 6708) D283 Moreover, the text, which is representative both for the subtype structure and each block, is only in some cases the one from Corpus A (the oldest one in the respective set, as the text D075 above) I 75 II 378 D266 s 19891 II 407 II 172 II 68 s 19891 II 407 II 172 051 I 118 C004 D225 6542 D262 6848 Some texts are the oldest representative ones for some blocks in the sequence but not all of them (as s 80 or C004). In such cases structure could be given priority over chronology. I 86 C033 C004 II 186 II 39 6848 (a)b: Leit g upp til himna / Sat g undir stofuvegg C082b I 67 060 016 6593 D303 6630 D200 s 80 I 125 II 104 D218 [sbr. D089] (f)K: g tti m r ... hest/hund + Kom g ar a kveldi II 190 D030 s 19550/19551 II 247 D078 s 19793 s 20124 D020 D305 s 19888 6570 KONA M N KOFANUM / KONA B NDA GEKK TIL s 2599 s 2245 s 20420 s 172 [endir] s 80 s 10935 D008 D052 D126 D283 D302 D234 s 2245 D234 s 80 II 334 D274 D165 D312 D008 D052 D165 s 21652 HVA ER R, BOLI MINN 6469 9330 D145 II 284 S KTU VATNI VIGGA ti stendur t kin 6462 6547 6430 8184 6708 6481 6567 6822 6851 [ 6600] 7501 II 293 I 64 s 257 D163 s 226 D123 Tartu, 6 July 2022: Plotting Poetry 5 Yelena Sesselja Helgad ttir: Textual Variation and Representative Selection of Texts 27

Results (preliminary!) Every major block in this sequence has a representative text albeit not necessarily the oldest one that is also representative for the structure of the respective subtype of the sequence, and each block of that respective sequence subtype The same is also true for ca. 75% of the subtypes of the major blocks NB the subtypes of the major blocks do not necessarily have a matching representative text in each subtype of other blocks the texts s 80, D052 and s 2245, representing three different subtypes of the block Kona m n kofanum, all belong to the same subtype of the block Sat g undir fiskihla a this also means that while one of the subtypes of the block Leit g upp til himna has three texts representative for that block, the respective sequence subtype and each block of that sequence subtype, the other two subtypes of that same block have none Tartu, 6 July 2022: Plotting Poetry 5 Yelena Sesselja Helgad ttir: Textual Variation and Representative Selection of Texts 28

Results (preliminary!) As to the questions mentioned above, i.e. whether the corpus resulting after all the blocks have been analyzed will also be representative of most PM block sequences, the answer is apparently positive judging from this one but large sequence the oldest representative texts of the blocks and their subtypes, which are part of this sequence, i.e. those texts that belong to Corpus A, also represent all the three main subtypes of the sequence updating Corpus A which already consists of the oldest representative text of each block with texts representing the sequence thus seems unnecessary I do not have numbers ready, but one and the same PM -text, composed of several blocks, is apparently quite often representative for more than one block Tartu, 6 July 2022: Plotting Poetry 5 Yelena Sesselja Helgad ttir: Textual Variation and Representative Selection of Texts 29

D075 D238 6570 6565 I 67 II 249 abeijkln prstvx (hann var m r a miklum au ) abcd {e/ }(h)ijklmo{p/q}rstvxy ( tti g a gj ta II 373 I 86 I 47 I 53 II 378 + abcd {e/ }(h)ijklmo{p/q}rstvxy + hann var m r a miklum au +) s 20116 s 19473 s 19785 s 19685 (II 28) II 249 s 19785 D075 D073 s 20385 s 19453 s 2572 I 39 C100 II 284 s 19421 s 19824 D150 s 20444 D225 D150 C038 045 I 94 C187 022 s 20116 7187-I 6570 6565 s 19453 D238 D231 II 75 II 225 I 39 I 67 I 94 II 59 II 75 II 137 F r g til ... / Gott tti m r t a l ta II 137 C002 C082a 016 (II 203)(C038) C002 C182 C038 C100 D075 D226 SOFA HLJ NAKORNIN B I 6919 (II 268) s 19880 6686 II 178 II 249 I 47 II 268 046 s 20444 s 19685 D231 022 s 20385 s 19880 s 19785 s 19824 I 125 II 53[d]II 104 II 172 II 407 LEIT G UPP TIL HIMNA KOM G AR A KVELDI g tti m r ... hest/hund SAT G UNDIR FISKIHLA A 060 C187 C004 D030 ab /ehijkl{mop}/{n p}rstvx (rak/r ku staf hnakka m r abcd ehijklmoprstvwy ( tti g a gj ta II 394 D030 s 19888 6565 C095 s 20124 6570 I 125 II 271 D070 s 20452 + ab /ehijkl{mop}/{n p}rstvx + hann var m r a miklum au ) abcd ehijklmoprstvwy + s an l h n steindau ) II 394 D262 II 225 I 86 (I2-D 83) C004 (g)h: F r g til sj var / Hva ger i go i s 21699 D008 6807 6427 II 238 046 II 271 s 19745 [og etv. fyrrv. -46?] II 238 II 31 C120 s 20566 D052 6893 6481 s 257[b] D114 6608 6463 s 2245D123 7137 7501 s 2277D126 7794 6547 I 64 D163 8411 6842 D190 7096 6600 s 10935 D201 6905 6451 II 293 D302 ( 6708) D283 I 75 II 378 D266 s 19891 II 407 II 172 II 68 s 19891 II 407 II 172 051 I 118 C004 D225 6542 D262 6848 I 86 C033 C004 II 186 II 39 6848 (a)b: Leit g upp til himna / Sat g undir stofuvegg C082b I 67 060 016 6593 D303 6630 D200 s 80 I 125 II 104 D218 [sbr. D089] (f)K: g tti m r ... hest/hund + Kom g ar a kveldi II 190 D030 s 19550/19551 II 247 D078 s 19793 s 20124 D020 D305 s 19888 6570 KONA M N KOFANUM / KONA B NDA GEKK TIL s 2599 s 2245 s 20420 s 172 [endir] s 80 s 10935 D008 D052 D126 D283 D302 D234 s 2245 D234 s 80 II 334 D274 D165 D312 D008 D052 D165 s 21652 HVA ER R, BOLI MINN 6469 9330 D145 II 284 S KTU VATNI VIGGA ti stendur t kin 6462 6547 6430 8184 6708 6481 6567 6822 6851 [ 6600] 7501 II 293 I 64 s 257 D163 s 226 D123 Tartu, 6 July 2022: Plotting Poetry 5 Yelena Sesselja Helgad ttir: Textual Variation and Representative Selection of Texts 30