The Greek Fragments of Thomas: Shared Sayings, Changed Contexts

The Bodleian record identifies P.Oxy. I 1 as MS. Gr. th. e. 7 (P), one papyrus leaf from a codex. Its date statement is third century with a question mark. The British Library's Papyrus 1531 is P.Oxy. IV 654: sayings on the reverse of a land-survey list. Harvard's Houghton Library holds P.Oxy. IV 655 as MS Gr SM4367, catalogued around 200–250, with six surviving fragments from a roll. [SRC-01] [SRC-02] [SRC-03]

These identifiers prevent a common visual mistake. The pieces are not three torn sections that we can simply assemble into one recovered book. They represent different copying occasions and physical formats. Their shared identification as Thomas concerns textual relationship. It does not establish a shared scribe, a single owner's library or a chain in which one of these surviving manuscripts was copied directly from another.

The current images are also distinct kinds of access. Oxford's two photographed sides and Harvard's first digital view were inspected here. The 1904 plates provide another representation of P654 and P655. Images allow us to notice fragment edges, missing areas and the disposition of writing, while an edition provides a reading argued from those traces. This article does not claim a fresh specialist transcription from the photographs or a physical examination of the originals. The difference between looking and editing should remain explicit even when both activities begin with the same image.

P654 makes the distinction unusually concrete. The British Library's broad record range covers the end of the second through the third century. Its detailed script fields distinguish the earlier cursive land list on the recto from the mid-to-late-third-century uncial sayings on the verso. Assigning the earlier writing's date to Thomas would collapse two acts of writing into one. [SRC-02]

The fuller Coptic Gospel of Thomas belongs to Nag Hammadi Codex II, a fourth-century witness now associated with the Coptic Museum in Cairo. It contains a collection conventionally divided into 114 sayings. [SRC-12] The Greek copies discussed here therefore offer earlier material witnesses to portions of a related text. That relative chronology is useful. It does not tell us when every saying originated, when sayings were first assembled, or when a particular arrangement became established.

Imagine that an old family letter survives only in a later copy. Dating the paper dates the copy, while the words may describe something earlier. That analogy does not settle Thomas's history; it clarifies which question is being asked. In this case, an ancient saying could have a history before the collection that contains it, and a collection could circulate in differing forms before the complete witness we possess. The appropriate conclusion is a set of limits and relationships, not an exact year extracted from a handwriting estimate.

Grenfell and Hunt's 1898 volume prints P1 with gaps and proposed supplements. Their 1904 volume does the same for P654 and P655, before the complete Coptic text was available to them. The original printed pages show why an edition is an argument as well as a convenient reading text: surviving letters, missing space and supplied language do not have the same status. [SRC-04] [SRC-05]

Later evidence can change that argument. The Coptic collection helped scholars recognize relationships and reconsider earlier supplements. Yet filling a Greek gap by translating Coptic back into Greek can become circular if the two witnesses actually differ. One first assumes agreement, uses that assumption to restore missing words, and then points to the restoration as proof of agreement. A responsible comparison keeps the conjecture visible and asks whether the proposed language fits the surviving traces and available space.

DeConick's 2006 paper illustrates the continuing work. After examining the originals, she challenges several restorations on grounds including the room available for letters. Her proposals are published scholarly judgments, not new ink appearing on the papyrus. [SRC-09] For a nonspecialist reader, the practical lesson is simple: retain brackets when presenting a reconstruction, and do not quote an uncertain completed sentence as though every word survived. The gaps have historical importance because they limit what the witness itself can establish.

P654's opening material includes the saying normally numbered Thomas 2. In the published Greek reconstruction, seeking leads through finding and astonishment to reigning and rest. The Coptic form includes disturbance or bewilderment and ends with reigning over all. Fitzmyer's side-by-side discussion makes the different endings visible; the Greek remains partly restored. [SRC-05] [SRC-06]

The difference is more than a choice between elegant English translations. It affects the movement of the saying. An ending in rest and an ending in reign over all invite different emphases, even while the sequence remains recognizably related. We should describe that relationship before deciding why it exists. It might involve a different textual form, editorial change or translation history; the surviving comparison alone does not identify the person responsible or the moment of change.

Matthew 7:7–8 also associates seeking with finding, a parallel already discussed by the first editors. [SRC-05] Shared language provides a starting point for comparison, not a complete family tree of sources. To argue that one text borrowed from another, one must do more than point to a familiar theme. Equally, to announce independence merely because a sentence has changed would overlook the possibility of adaptation. This short example teaches us to separate recognizing a parallel from reconstructing the entire route by which a saying reached the page.

P1 places the saying about finding Jesus when lifting a stone and splitting wood immediately after language about being with a person who is alone. In the Coptic collection, the stone-and-wood material appears later, within saying 77, after a declaration involving light and the all. Even the order of stone and wood is reversed. The Greek printed text and the Coptic comparison make this change of setting clear. [SRC-04] [SRC-06]

Placement can matter as much as the retention of a memorable phrase. The same words heard after a promise of companionship may strike a reader differently when attached to a cosmic declaration. That observation does not prove which setting came first in the history of composition. Nor does it license us to turn either version into a precise doctrinal formula without interpreting its language. It shows that comparing isolated quotations can conceal the very differences that make the witnesses valuable.

A fair presentation should therefore give the reader the neighboring thought as well as the attractive line. We can acknowledge that the sayings are related without describing the Coptic manuscript as a mechanical translation of this particular Greek leaf. We can also leave the difficult, damaged language earlier in P1's same passage open to further specialist work. The relatively clear placement of the stone-and-wood saying does not resolve every disputed word beside it. Historical understanding grows by retaining that unevenness of confidence.

The material corresponding to Thomas 36 in P655 is longer than its Coptic counterpart. The Greek edition includes language about food, clothing and lilies, with several restorations; the Coptic saying is much shorter. The Greek also recalls Matthew 6 and Luke 12, although it does not reproduce either passage word for word. [SRC-05] [SRC-06]

Here length supplies a useful caution. An earlier surviving copy can preserve a longer form. Therefore shorter cannot simply mean earlier, and later cannot simply mean expanded. We must examine the actual relationship at each point. An abbreviated Coptic form, a fuller Greek form, different exemplars and stages of collection remain possibilities to assess; the existence of extra material is not by itself a confession of deliberate deception.

The first publication's plate also records a particular arrangement of fragments. Harvard's present description notes that two small pieces from the original eight are no longer present. [SRC-03] A clean modern paragraph can hide that physical situation. The printed reconstruction is valuable precisely when it continues to show how much of the sentence is supplied and how the surviving pieces constrain it. Readers need not master every letter to understand the methodological consequence: a comparison between reconstructed passages requires greater care than comparing two complete printed pages.

Hurtado's study considers book form, writing and reading aids, favoring private or nonliturgical use for these copies while retaining limits. Luijendijk argues that P1 could have been read in worship and explores how features of the other manuscripts complicate a rigid private/public division. These are differing inferences from material and literary evidence, not recovered attendance lists. [SRC-07] [SRC-08]

A codex does not announce its liturgical status merely by having pages. Nor does reused writing material identify its owner as poor, careless or doctrinally marginal. Format, handwriting and punctuation are meaningful clues when considered with comparisons and context; they are not one-step tests for canon or community identity. The disagreement matters because it prevents a modern reader from treating a plausible reconstruction of use as a documented event.

We can still reach a substantial conclusion: people copied and engaged with forms of these sayings. The manuscripts deserve a place in the history of Christian reading. From there, claims about precisely who owned them, whether a named bishop heard them, or what every worshipper regarded as Scripture require further evidence. A Catholic account should not erase unfamiliar ancient reading practices. It should also avoid inventing a suppressed community whose entire life is supposedly recoverable from the shape of three books.

Calling the collection the Gospel of Thomas identifies the work under discussion. It does not establish that the apostle composed it. Benedict XVI explicitly speaks of apocryphal works written in Thomas's name. [SRC-10] That wording allows us to acknowledge the ancient attribution without converting the attribution into an independently verified author biography.

The collection should also be distinguished from the Infancy Gospel of Thomas, a different work presenting stories about Jesus as a child. The latter's surviving narratives include incidents with other children, teachers and Joseph. [SRC-13] A claim about a newly examined infancy fragment cannot automatically be applied to the sayings manuscripts here. The shared name can make search results confusing; the text, shelfmark and contents must do the identifying.

For Catholic reception, Dei Verbum 18–19 identifies the fourfold Gospel according to Matthew, Mark, Luke and John as the Church's apostolic witness to Jesus. [SRC-11] The historical survival of an additional ancient book is not an additional act of canonical reception. Conversely, a noncanonical text's historical interest does not disappear because it lacks that authority. Keeping these distinctions clear permits rigorous comparison without asking every papyrus discovery to reopen the whole canon or pretending that study amounts to ecclesiastical approval.

The assessment is VERIFIED for a bounded proposition: these three Greek witnesses preserve sayings related to the later Coptic Thomas, with observable differences in wording or arrangement. That conclusion survives disagreements about exact restorations and early reading settings. It does not establish that every saying goes back to Jesus, that the apostle wrote the collection, or that one surviving manuscript represents an unaltered first edition.

Nor does this article settle the larger scholarly debate about Thomas's relationship to the canonical Gospels. The worked comparisons establish parallels and differences at particular places. A comprehensive dependence argument would require a much wider assessment of the collection, languages, sequence and sources. It would be misleading to announce a verdict on that whole debate after examining only selected passages, however informative those passages may be.

The access record matters to that limit. Original early editions, custodian descriptions, selected manuscript images and the disclosed critical discussions were examined. The complete Coptic manuscript was not independently transcribed or collated, and not every later critical edition was obtained. Human verification remains pending. These limitations do not cancel what the inspected material shows; they describe the further work needed before making stronger claims. An honest article can leave that work visible and still give the reader a useful, concrete understanding of what has survived.

There is a difference between collecting striking sayings and allowing Christ to form a life. A memorable sentence can comfort us while escaping its context; careful reading asks how it relates to the speaker, the surrounding passage and the way we act. The comparisons above can sharpen that attention when we return to the canonical Gospels. They need not become a contest over which book offers the most surprising quotation.

Try reading a short passage from Matthew or Luke twice. On the first reading, notice the sentence you instinctively want to share. On the second, notice what comes immediately before and after it. Does that setting challenge the way you first understood it? Then consider one practical response: a truthful apology, a generous action, a moment of prayer that you have repeatedly postponed. This exercise is an original Catholic application, not a reconstruction of an ancient Thomas reader's devotion.

Wonder can survive the admission that a manuscript is incomplete. Indeed, the fragile witnesses make the labor of listening more tangible. Unknown people copied these words; others preserved, studied and sometimes corrected their readings. We can be grateful for that history without borrowing certainty from it. Bring the questions into prayer with a willingness to learn, and bring the conclusions into conversation with enough precision that another person can distinguish what the evidence says from what you hope or believe. Such care is one way curiosity becomes service.