Gripla - 2020, Síða 23

GRIPLA22
whereas no differences were noticeable between CX [C 13–18,
C-divergent] and I–II [C 1–4, 19–21, the parallel chapters].48
Thus Erichsen argues that the A-redaction is likely a result of the mixing
of sources, whereas the C-redaction is more internally consistent. In the
stylometric analysis that follows, we agree with this conclusion.
Magerøy, on the other hand, argues that the primacy of the A-redaction
is reflected syntactically when compared to C-divergent, but also with
the C-redaction’s þættir standing out in particular. For example, he finds
a higher use of “ok” in chapters 1–4 and A-divergent compared with
C-divergent;49 he repeats Björn Sigfússon’s assertion that there are more
dependent clauses in C-divergent than in A-divergent and chapters 1–4,
noting an exceptionally high number of these clauses in the C-redaction
þættir; also, like Erichsen, he agrees that C-divergent is characterized by
hypotaxis and A-divergent by parataxis, but unlike her he suggests that
chapters 1–4 are characterized more by hypotaxis than parataxis.50
At this point the stylistic discussion largely ended with the matter re-
maining unsettled. With the Freeprose-Bookprose debate losing steam as
the twentieth century went on, so too did the Ljósvetninga saga redaction
problem drift away from scholarly attention. The task of this study is to
engage with the problem once again, leveraging the advent of stylometrics.
48 “Hier muss man eine Quellenmischung, innerhalb der schriftlichen Textüberlieferung
annehmen: ein Schreiber — sei es der von 162 oder einer Vorlage, sei es der von 561 oder
einer Vorlage — hat dieses Mittelstück aus einer Nebenquelle (einer mündlichen oder einer
schriftlichen) geschöpft, vermutlich weil die Hauptvorlage hier eine Lücke hatte, oder auch
weil ihm gerade für diese Strecke eine Quelle zu Gebot stand, die ihm besser gefiel... Mit
anderen Worten: A ist das Ergebnis einer Mischung, und C folgt einer zusammenhängen-
den Vorlage, oder umgekehrt. Von diesen zwei Möglichkeiten ist die erste vorzuziehen;
denn Wortschatz und Stil von AX weisen immerhin einige Züge auf, die von I und II (in A
wie C) abweichen wogegen zwischen CX und I–II keine Unterschiede bemerkbar [sind]
…” Erichsen, Untersuchungen, 59–60.
49 Magerøy here uses the 561 readings for the parallel chapters 1–4 and 19–21 for comparison
with both A-divergent and C-divergent. This assumption does not take into account
manuscript transmission. If Magerøy had counted the use of ‘ok’ in C chapters 1–4 and
19–21, he would have found that ‘ok’ is a scribal tendency in 561 and not necessarily a styl-
istic feature of A-redaction, discussed further below.
50 Note that he frequently reveals that the gap between C-divergent and A-parallel widens
when the parts of chapters 13–18 that are not extant in A-divergent are taken into account.
However, it could very well be that dependent clauses are a stylistic characteristic enforced
by the plot itself, for example due to the introductory nature of these chapters.