The textual tradition of the Lancelot-Grail, and especially its central part, the Lancelot proper, is complex. Multiple versions exist of all constituent parts, with scholarship usually distinguishing between a long version and a shorter version, or an abridged one in the case of the Queste and Mort. Recently, Richard Trachsler has argued that in the case of the Suite du Merlin the different versions may be connected with their relation to the rest of the cycle. To make matters even more complicated, transcriptions tend to 'shift' between the established versions, some just once or a few times over the course of the entire text, others more frequently. These latter copies are usually indicated as 'oscillating' MSS. The following exposition will be limited to the Lancelot proper, and may serve to elucidate the choices that were made in the segmentation of the text.
In a series of articles published in Romania between 1964 and 1966, Alexandre Micha analysed the complicated textual tradition of the Lancelot proper. His findings were later further developed in the introductions to his edition of the Lancelot en prose and a number of more detailed investigations, the most influential of which were collected in his Essais on the Grail cycle (Micha 1987). Notwithstanding that it was possible to distinguish between four stable groups of MSS (with sigla a, d, e and i), which in yet other copies appeared in different mixed constellations, Micha found that, in general, the text of the first part of the Lancelot – recording Lancelot's childhood and installation as one of Arthur's knights – was common to all extant MSS (Micha, Ia - LXXIa).
From that point onwards things become more confused. A group of five MSS (Paris BnF f.fr. 768, Rouen BM 1055, Florence BML Pluteo 89 inf. 61, New York Morgan 805-6, and Paris BnF f. fr. 339) give a 'special' version (in Micha’s terminology also 'version B') of the journey to Sorelois and the False Guenevere episode (Micha, I-XXXV). Notwithstanding that the latter two MSS continued the narrative beyond the boundaries of this diverging text, it appeared that this version was meant to end with the death of Galehaut, the giant king of the Distant Isles. The introduction of this non-cyclic version (Micha I – II,9 ) was also preserved in a further group of seven MSS (Paris BnF f.fr. 110, Paris BnF f. fr. 111, Paris BnF f.fr 112, Paris BnF f.fr. 114, Bonn ULB S 526, Grenoble BM 865, and London BL Add. 10293), which later continued with the regular text (which Micha referred to as A). That BnF f.fr. 768, dated to c. 1220, and the oldest extant witnesses of the cyclic Lancelot propre are (near-)contemporary, makes it especially difficult to establish the order in which these versions were composed. A number of compositional specifics of this shorter version, for instance the simplification of narrative structure, chronology and reference to an extra-diegetic Arthurian universe, led Micha to believe that a later remanieur had abridged the longer, cyclic account which continued with the Charette episode, its Suites and the Agravain. This position directly opposed that of Elspeth Kennedy, who was convinced that the shorter text, edited in Kennedy 1980, presented a coherent narrative centred around the themes of identity and love, and concluded with Lancelot being rewarded for his merit. In this version, which in Kennedy's view had preceded the longer, cyclic one, Perceval and not Galahad had been destined to accomplish the Grail Quest.
According to Micha, the MSS of the cyclic text, i.e. those continuing the narrative beyond Galehaut's death, could be further divided in a short version (also indicated as C or London, named after BL Add. 10293) and a long version (L or Paris, named after Paris BnF f.fr. 344). In this group of MSS as well, mixed versions occurred. The most important of these is the significant group of some eleven copies. Right before the episode Micha XXIV,35, the text in these MSS shifts from the long cyclic version to the short cyclic version. These so-called E-type MSS have been named after the siglum of Paris BnF f.fr 1430, one of the oldest extant members of the group, which may be dated as early as 1230-50, thus providing a terminus a quo for the cross-pollination of the long and short versions. Given that the most ancient witnesses of all versions are dated to approximately the same period, one cannot confidently state which version was first. Again, it was the process of rewriting, for instance the omission of important narrative elements leading to contradictions in the story, which suggested to Micha that the shorter version was probably a curtailment of the longer narrative, possibly inspired by commercial considerations.
In the Charette episode (Micha XXXVI-XLII), which tells the story of Lancelot's humiliation before Guenevere, two groups of MSS continue to oppose each other. However, according to Micha, the process of rewriting was different, and also the distribution of copies did not completely confirm to the division between long and short in the previous sections. Because of these considerations, he decided to refer to the versions in the Charette as 'alpha' and 'beta', rather than 'long' and 'short' as in the other sections of the text. Aditionally, Grenoble BM 865, together with the eleven E-type MSS (together referred to as version betabeta), presented a revised introduction, which differed from the larger beta-group, to which these MSS otherwise belonged. This introduction bears close resemblance to that of a final group of three MSS (BnF f.fr. 112, Paris BnF f.fr. 119 and Paris Arsenal 3480), substituting the Vulgate version of the Charette with a novel de-rhymed version of Chrétien's verse romance, combined with parts from the Vulgate redaction, indicated as 'version gamma' (ed. Combes 2009). After the Charette, in the so-called Suites (Micha, XLIII-LXIX) the MS versions return to their original constellation of long versus short; and in the final part of the text (Micha, LXX-CVIII), usually referred to as the Agravain, matters again become slightly more complex, when different mixed and oscillating redactions appear.
However complicated the tradition may seem, the above only scratches the surface of the actual complexity of the textual transmission as evidenced by the extant MS record. In a seminal article, Elspeth Kennedy emphasized the importance of scribal authorship and the mechanics of MS production in the genesis of textual variation (Kennedy 1970). The scribe of London BL Royal 19 C XIII has significantly abbreviated the text, in order to produce a copy at lower cost or otherwise to lessen the workload. Another striking example may be found in Rouen BM 1054, where the scribe has left out certain episodes of the first quest for Lancelot at the explicit request of the lady he is working for: she was apparently primarily interested in the Charette. Other MSS have additional, newly composed episodes, for instance to underline Gauvain's prowess. In our database, these episodes have been marked as interstitial segments (see: Segmentation), since in this event they may assist in identifying closely related copies.