Most scribes left no records of themselves except the manuscripts theywrote. Some, however, left their names and other information in the colophonsof the manuscripts they wrote. Colophons -- a scribe's "signature"of his manuscript -- are almost unknown in early documents, but becomerelatively normal in late minuscules.
The name "colophon" is relatively recent.Erasmus concluded many of his writings withthe phrase colophonem addidi, and the phrase came to be applied toanything at the end of a work. But the principle of a brief note at the endof a copied work is very much older.
Colophons could contain almost anything: The date of the manuscript(usually in the form of the Year of the Worldand/or the indiction), the scribe whowrote the manuscript, the type of manuscript it was copied from, the placeit was copied, or the person for whom it was copied. The date on which amanuscript was copied is always useful, of course. But it can also beuseful to know where it came from (since it allows us to say thata certain sort of text was in circulation there at a certain time). Knowing a scribe'sname is also interesting, though it really doesn't matter much unless we haveother works from his pen.
Colophons could also contain various petitions and requests (e.g. a prayerfor God to forgive the scribe or a request for a reader to take good care ofthe copy), but these have little importance except, perhaps, as a source ofinformation about the liturgical usage of the time. The colophon in S(the first and only uncial to have an intact colophon, though we find earlierscribal signatures, e.g., in the minuscule 461 and in the Latin CodexFuldensis) is not atypical:
εγραφει η τιμιαδελτος αυτη διαχειρος εμουΜιχαηλ μοναχουαμαρτωλουμηνι μαριωα α'.ημερα ε',ωρα ς',ετους συνζ. ινδ ζ'-- i.e. it is the work of "a monk, a sinner"named Michael who finished his task in the sixth hour of the fifth day ofMarch in the year 6457 (949 C. E.).
The subscription to the Pauline Epistles in 1739is not all that different; although it omits the date (possibly given in oneof the excised portions of the codex, as each part had a colophon), ittoo gives the scribe's name(Εφραιμ μοναχου)and begs Godfor mercy. Elsewhere in 1739, Ephraim gives us information about how hismanuscript was compiled.
There seems to be a certain tendency for colophons to grow more elaborateover time, though of course they continue to be highly individual.
Interestingly, not all colophons are accurate; some are forgeries.Colwell, in "Method in Validating Byzantine Date-Colophons:A Study of Athos, Laura B.26" (now available in Colwell's Studies inMethodology in New Testament Textual Criticism, pp. 142-147) offers thecase of manuscript 1505,which has a forged date of 1084 (note: letterforms are modernized andthe line breaks of the original are not retained):
Taking the first two items first, we see that the manuscript datesitself to the reign of the Byzantine Emperor Alexius Comnenus (reigned 1081-1118),and specifically the year 6592 (=1084 C. E.). However, the remainingdata (sun cycle, moon cycle, indiction, Sunday of abstinence from meat, legalpassover, Christian passover, and fast of the holy apostles) do not correspondto 1084, and indeed other colophons from the eleventh century often do not evenlist most of these last, which are typical of the fuller colophons of aboutthe fourteenth century. The data appears to correspond, in fact, to the year1445. As the colophon is not in the same hand as the rest of the manuscript(which would appear to date from the twelfth century), it seems clear that itwas forged to make the manuscript appear older and more valuable (though,interestingly, the colophon makes it only slightly older than what seems tobe its actual date, and since 1505 belongs to Family 2138, its basic text is infact older than the colophon suggests). Colwell cites other instancesof this sort of forgery. Therefore even colophons must be treated with somecare.
We also seem to have instances of scribes forging names. 223has a colophon attributing it to Antonios of Malaka (who is also associatedwith 1305 and 
279) --but the colophon to 223 is not by the same hand as the manuscript, and theother two Antonios manuscripts are dated 1244 and XII, respectively, while223 appears to be from the fourteenth century.
This is not confined to Greek manuscripts. The Book of Dimmais a small Irish copy of the Vulgate Gospels (Trinith College, Dublin, MS. A.4.23 (59)).It claims to have been written by Dimma Mac Nathi. There is a mention in theLife of St. Cronan that a Dimma wrote a gospels book in the life of thatsaint, the founder of Roscrea monastery. Cronan died in 619, so the colophonseems to be trying to make the book both older and more significant than it is.(It would be interesting to know if the scribe tried to put it in the Roscrealibrary with that inscription, or if it was destined to be sold or given away,with the scribe trying to increase its value.)
In some cases it is quite interesting to know the several manuscriptsfrom a scribe's pen. This is true, e.g., of Ephraem, who gave us two of themost important of all minuscule manuscripts (1582 and1739), plus texts ofAristotle and Polybius. We also observe that manuscripts from the same scribeare often akin textually (observe the Kx Cluster 74 manuscriptswritten by Theodore of Hagiopetros; these represent a third of the manuscriptsof this type. Even more extreme is the case of George Hermonymos, who wrote atleast five of the seven manuscripts of Kx Cluster 17).
On the other hand, sometimes a scribe would copy manuscripts of very different types. We don't know the name of the scribe of 1505, mentioned above -- but we believe that he wrote at least three manscripts, 38, 1505, and 2400, all of which contain the Gospels, Acts, and Epistles. Yet 1505 has a Family 2138 text of the Acts and Epistles, 38 seems to have a weak form of the Family 1319 text, and 2400 is just an ordinary Byzantine text. Clearly this scribe, at least, had no clear preference for a particular type of text.
It is unfortunate to note that some monasteries discouraged scribes from including their names in colophons (including one's name was seen as a mark of pride). This not only makes it harder to identify manuscripts from the same copyist, but discourages the inclusion of other useful information in colophons. Fortunately, not every scribe paid attention. And even a colophon without a name or date can often help us date a manuscript. This can happen, e.g., if the scribe uses datable words or phrases. This is more typical of secular works (an example is the Peterborough manuscript of the Anglo-Saxon chronicle, which has been interpolated with stories localized to that area; we learn much from the language used in the insertions), but it can apply to Biblical manuscripts too if the colophon is long enough.
The table below lists certain of the scribes known to have writtenNew Testament manuscripts, along with the manuscripts copied and theirtext-types as far as known (Gospels classifications are generally from Wisse,unless marked VS: for Von Soden; other descriptions are from Von Soden or thepresent writer). Note: Some manuscripts are identified with particular scribesonly by the handwriting; no attempt is made to distinguish these.
After each scribe's name, in square brackets [], are the dates at whichthe manuscripts ascribed to him were written (based on the colophons orpaleography).
| Scribe | Manuscripts |
| Abraham Teudatus [XI] | 507 (Kx) |
| Andreas [1111] | 203 (VS: ap: Ic2, r: K) |
| Andreas [XI/XII?] | 180 (in gospels; John added the rest of NT later; Kx Cl 180) |
| Angelo Vergèce [XVI] | 296 (VS: e: Kx, apc: Ib1, r: Ia2), 1931 (VS: Ia) |
| Anthony [XI] | 343 (Cl 343/Kmix) |
| Anthony [1506] | 445 (VS: Kx) |
| Arsenius [XII] | 862 (VS: Θε29) |
| Athanasius [1434] | 616 (VS: Ic?) |
| Basil Argyropolus [1140] | 229 (Πa/Kx) |
| Calistus [1432] | 286 (Kx) |
| Constantine [1052] | 174 (Λ) |
| Constantine [1326] | 492 (Kx) |
| Constantine Chrysographus [XII] | 347 (Kx) |
| Cosmas Vanaretus [XIII] | 503 (VS: Kx) |
| Dionysus [XI] | 506 (e: Cl 276; VS: ap: Ic2 r: K) |
| Ephraem [949, 954, X] | 1582 (Family 1), 1739 (Family 1739), Cod. Marcianus 201 (of Aristotle's Organon, at Venice; dated 954), Cod. Vat. gr. 124 (the leading manuscript of Polybius, probably to be dated to 947) |
| Euphemius [1043] | 609 (Greek/Arabic; M609), |
| Eustathius [XII] | 129 (Kx) |
| George Hermonymos of Sparta [1478, XV] | 17 (Kx Cl 17), 30 and 30abs (30 is Kx Cl 17 with 288), 70 (Kx Cl 17), 287 (Kx Cl 17), 288 (Kx Cl 17 with 30), 1848 (VS: Kc) |
| George [XIII] | 579 (B) |
| George [XIII/XIV] | 429 (George is responsible only for the Acts and Epistles; r is from another hand. VS: ap: Ib1; r: K; despite von Soden, it is easily shown that 429 is part of the group 206-429-522, which is goes with Family 1739 in Acts and Family 2138 in the Catholics.) |
| George [1305?] | 649 (VS: Θε408) |
| Gabriel [XV] | 525 (Greek/Slavonic, with the Greek later and probably by an anonymous hand; Kmix/Kx/TR) |
| Gerasimus [XIV] | 498 (e: M1386 ap: VS: Kr) |
| Gregory [XII] | 438 (Kx) |
| James of Sinai 1316] | 489 (e: Πa with 1219; ap: VS: Ia2) |
| Joachim, George, and others [XII-XIV?] | 632 (VS: p: K) |
| Joasaph [XIII] | 410 (M349) |
| Joasaph [1366, 1369, 1376, 1394] | 480 (Kr), 634 (VS: Kr), 1100 (VS: Kr), 1960 (not classified by Von Soden or Aland/Aland; seems to have at least some Kr readings) |
| John [1044] | 81 (VS: H) |
| John (of Patmos) [XI] | 1194 (M10) |
| John [1179] | 688 (Kx Cl Ω) |
| John [1199] | 245 (Kmix/1167) |
| John [XII/XIII] | 421 (VS: K) |
| John [1273] | 180 (in Acts, etc.; written by Andreas in the Gospels) |
| John Rhosus of Crete [1478] | 448 (Kx Cl 183) |
| John Serbopulos [XV] | 47 (Mix/Kr), 56 (Kr) |
| John Trithemius [XV] | 96 (VS: Kx) |
| John Tzutzuna [1092] | 459 (VS: ap: H? r: Ib2) |
| Joseph [XI] | 422 (Kmix/Kx; John probably from another hand) |
| Leo [1039] | 164 (Λ with 1443) |
| Leo [XII] | 502 (Kx Cl 74) |
| Leo [1330] | 425 (VS: K) |
| Leontius [XI] | 186 (VS: Ac) |
| Lucas [1625] | 289 (VS: Kx) |
| Manuel [1153] | 162 (Kx/Kmix) |
| Manuel [1262] | 293 (M1195) |
| Maurus [XIII] | 427 (Mix/Kx/Kmix) |
| Meletius [1275] | 248 (Kmix/M27) |
| Michael [949] | S/028 (Kx Cl Ω) |
| Michael [1330] | 394 (e: Kr Gr 35) |
| Michael Damascenus [1515] | 522 (VS e: Kx; ap: Ib1, r: Ib; in fact part of the group 206-429-522, which is Family 1739 in Acts and Family 2138 in the Catholics.) |
| Neophytus [1305] | 645 (Kr) |
| Nepho [1159] | 439 (Kx with 877) |
| Nicephorus [1092] | 276 (Cl 276) |
| Nicetas Mauron [1296] | 341 (VS: Kx) |
| Nicholas [835] | 461 (Kx Cl Ω) |
| Papadopoulous Kerameus [1344] | 1766 (VS: Kc) |
| Paul [XI] | 26 (Kmix/Kx) |
| Philip [XIV] | 414 (M349) |
| Philotheus [1314] | 235 (Kmix/Kx) |
| Synesius 1033] | 504 (Kx) |
| Theodore of Hagiopetros [1278, 1280, 1284, 1292, 1295, 1301] | 74 (Kx Cl 74), 234 (Kx Cl 74), 412 (Kx with 1394), 483 (e: Kx Cl 74; ap: VS: Kc), 484 (Kx Cl 74), 856 (Cl 2148), 1594 (Kx Cl 74) |
| Theodore [1037] | 623 (VS: Ia2; Richards: Family 1739, but with too low a percentage to be meaningful) |
| Theodosius [1338] | 54 (Kmix/Kx) |
| Theodosius ρακενδυτης [1302] | 413 (Kx Cl 143) |
| Theophilus [1285] | 482 (Kx/Πa) |
| Theophylact [984] | 619 (not classified by Von Soden or Aland/Aland) |
Even when a scribe does not need a colophon, we can often tell somethingabout him beyond his approximate date. Letterforms, artwork, marginal equipment --all can tell something about the scribe. An obvious example is Irish scribes. RobinFlower wrote of these, "Irish scribes -- and only Irish scribes [during theninth century] -- had a habit of setting down in the margins and on blank spacesof their manuscripts personal memoranda, invocations of saints, little fragmentsof verse, and all the flotsam and jetsam of idle fancy" (Robin Flower, TheIrish Tradition, [1947], p. 36). Flower's examples are mostly from non-Biblicalmanuscripts, but there is a well-known example in Codex Boernerianus(Gp) of a scribbled note, in Gaelic,regarding a pilgrimage to Rome. This may not be from the original scribe, but otherexamples are.
Sadly, New Testament critics seem to make little use of the peculiarities ofscribes. Many scribes had peculiar spellings (e.g. both D/06 and 462 have problemswith -ε versus-αι. In the sections Ichecked, 462 has not a single verb ending in-ε; all had beenchanged to end in -αι).Obviously such manuscripts are useless for variants involving such verbendings. But such peculiarities may also tell us something about thenationality or dialect of the scribe, or the school in which he was trained.
We also know, e.g., that the chief peculiarity of the scribe of P75was omitting short words.
Useless information? Hardly! Shakespearean scholars write whole theses aboutthe peculiarities of the typesetters who set individual pages of his works.Although this is partly of necessity (they have nothing else to work on), theamount of information they gain is simply astonishing. New Testament scholarscould surely derive many of the same benefits -- but it's a rare discussion ofa reading which makes any reference to scribal habits. It's a clear lack.
This is not the place for a long list of such peculiarities (since I have notthe data to compile such a list), but knowledge of such features belongs inevery paleographer's toolkit, and such peculiarities should be noted in editionsof manuscripts.
Incidentally, we know more about scribes than just what they tell us incolophons. There are several sources of information: The manuscripts themselves,the artifacts left by scribes, and illustrations in the manuscripts. Very manymanuscripts, e.g., contain illustrations thought to be the evangelists. Whatis noteworthy about these is how often they contain the same sort of image:The evangelist sitting before a codex (not a scroll!), with a pen in the righthand and a knife in the left. The illustration at right, from a late Latinmanuscript of John, is quite typical. The knife is not just used to sharpenthe pen; it is kept always in the left hand, presumably to scrape off mistakesbefore the ink can dry.