The central conclusion of The Myth of Clement is that the surviving writings attributed to Clement cannot support an invariant picture of his scriptural text, citation habits, or interpretive behavior. The documents contain a range of practices. That range has to be established before “Clement” can be used as a standard for deciding the authorship of a disputed passage.
This matters for the Secret Mark debate because an argument can appear decisive when its comparison figure has been made artificially consistent. A word, a Gospel formulation, or an interpretation is declared unlike Clement. Yet the accepted writings may already contain an equivalent degree or kind of variation. The first task is therefore to test the comparison figure itself.
The study changes the basis on which a claim of incompatibility with Clement must be argued. It documents the behavior against which that claim has to be tested. The following are the conclusions I draw from the comparisons.
1. Clement cannot be treated as a fixed textual benchmark
The study assembled 1,488 Gospel and Pauline reference records and identified 215 biblical verse coordinates represented in more than one work or collection. Of those comparison groups, 54 were classified as showing different wording, 84 as abbreviation or paraphrase, and 31 as agreement in the quoted overlap. The remainder involved insufficient Greek, range references, or allusive material. One further comparison examined corresponding Synoptic sayings under separately identified coordinates.
Those figures describe reference groups. Several coordinates can belong to the same ancient quotation, and a reference to a range does not reproduce every verse in that range. They are useful for organizing the evidence, rather than counting independent witnesses or assigning every difference to a separate manuscript.
The secure conclusion rests on the actual passages: the accepted comparison corpus already contains materially different formulations of familiar scriptural sayings. Some changes are small. Others alter a divine designation, the object of knowledge, or the parties addressed by a command. The differences occur within continuous works as well as across collections.
Clement can have recognizable tendencies while displaying this range of behavior. An attribution argument needs to establish how characteristic a feature is in comparable material. It cannot turn a tendency into an exceptionless rule simply by calling it Clementine.
2. The clearest differences concern the words actually cited
Three examples make the point especially clearly:
| Passage | Forms presented in the study | Immediate conclusion |
|---|---|---|
| The good-God saying: Mark 10:18 and its Synoptic parallels | Quis dives salvetur 4.5 has “the one God.” Paedagogus 1.8.72.2 has “my Father in the heavens.” | The corresponding saying is presented with different divine designations in two continuous works. |
| Ephesians 4:13 | Stromata 4.21.132.1 has “knowledge of the Son of God.” Paedagogus 1.5.18.3 has “knowledge of God.” | The object of knowledge differs within a closely comparable Pauline quotation. |
| Romans 14:3 | Paedagogus 2.1.10.3 tells the eater not to despise the abstainer. In Stromata 3.6.52.3, the abstainer is told not to despise the eater. The judging clause is correspondingly reversed. | The supplied texts change the assignment of the two commands. |
These are changes in the cited formulations, rather than merely different modern biblical references attached to the same words. They require an explanation even when the larger argument remains compatible across the passages.
Other examples establish the surrounding range. The Father–Son revelation saying appears with ἔγνω, “knew,” and ἐπιγινώσκει, “knows.” Kingdom sayings alternate between God and heaven. The Word is described as with God in one Johannine formulation and in God in another. Pauline exhortations acquire changed titles, pronouns, grammatical forms, or explanatory words.
The cause of a difference remains a separate question. Memory, adaptation, combination of passages, different source texts, and later transmission are possible explanations. The observed difference is established before a particular explanation for it has been demonstrated. Neither a reconstructed lost Gospel nor a theory of authorial carelessness should be inserted automatically at that second stage.
Sources: Quis dives 4.5; Paedagogus 1.8.72.2; Stromata 4.21.132.1; Paedagogus 1.5.18.3; Stromata 3.6.52.3; Paedagogus 2.1.10.3.
3. Verbal consistency and theological consistency are different questions
The study gives especially clear evidence against a fixed citation profile. Its implications for belief require a different kind of analysis. Changed words do not necessarily amount to changed beliefs.
Both versions of the eater-and-abstainer command occur in arguments resisting condemnation based on food. The practical concern can remain recognizable while the subjects of the clauses change places. Similarly, the God/Father difference in the good-God saying does not require Clement to be affirming two opposed divine beings. In the Paedagogus, the saying belongs to an argument identifying the good Father with the Creator.
There are also positive controls. The holy, just, and good commandment of Romans 7:12 recurs closely in the Stromata and Paedagogus. Other short Pauline clauses show close agreement. A comparison that retains those agreements can distinguish a stable formulation from an adapted one.
The resulting picture is more precise than either universal consistency or universal contradiction. A recurring theological purpose can coexist with different quoted wording. A theological profile therefore cannot, by itself, establish the exact text Clement must have quoted. Conversely, a verbal variation cannot, by itself, establish a new theological position.
This distinction matters whenever a disputed sentence is said to contain the wrong kind of Clementine theology. The argument must first establish what the sentence is doing: quoting a source, describing someone else’s position, adapting a saying, or expressing the writer’s own conclusion.
4. The Clementine corpus preserves more than one speaking voice
Some apparent incompatibilities arise because different voices have been combined under an author’s name. The Stromata reports opponents before answering them. Excerpta ex Theodoto preserves mixed material and sometimes explicitly identifies Theodotus or followers of Valentinus. Fragments may come through later quotation, commentary, or catena transmission.
The adversary saying provides a decisive example of the problem. Stromata 4.14.95.2 identifies the adversary as the devil and expressly rejects the body interpretation. Excerpta 3.52.1 calls the flesh the adversary. The interpretations oppose one another. Their opposition does not authorize us to assign both positions, without qualification, to Clement personally.
The name attached to a corpus does not identify the speaker of every statement inside it. An account of Clement’s beliefs must preserve the distinctions between authorial explanation, cited Scripture, an opponent’s argument, an excerpted teacher, and a later witness’s report.
The same issue applies to the Letter to Theodore. Its attributed speaker, the Gospel passages it quotes, and the opponents it describes are distinct objects of analysis. The accepted comparison material already requires those distinctions. They should be applied consistently when the disputed letter is examined.
Sources: Stromata 4.14.95.2; Excerpta ex Theodoto 3.52.1.
5. A Gospel title does not guarantee a modern canonical wording
The rich-man quotation in Quis dives salvetur is particularly important. Clement gives an extended passage and identifies it as written in the Gospel according to Mark. Yet the quoted passage includes:
εἰ θέλεις τέλειος εἶναι
“if you wish to be perfect”
That is the perfection formula associated with Matthew 19:21. The passage also retains the distinctively Markan statement that Jesus looked at the man and loved him. Clement’s own Mark attribution thus accompanies wording that requires comparison across Gospel forms.
The conclusion is that the title and the cited words must both be examined. Replacing the ancient quotation with the text of a modern edition would remove part of the evidence. The alternatives include a different textual form, remembered combination, or adaptation; the source label alone does not decide among them.
Clement’s Gospel knowledge also extends beyond material indexed under four canonical titles. In Stromata 3.13.93.1, he expressly distinguishes a saying absent from the four Gospels handed down to his community and locates it in the Gospel according to the Egyptians. The statement preserves both knowledge of an additional Gospel source and a distinction between that source and the four transmitted Gospels.
That evidence establishes a broader literary setting. It does not establish equal authority for every Gospel Clement discusses, or identify his rich-man quotation with the longer Gospel described in the Letter to Theodore. Those are further historical claims requiring their own evidence.
Sources: Quis dives salvetur 4.4–5.1; Stromata 3.13.92.2–93.3.
6. An “un-Clementine” feature has to be demonstrated against the observed range
A difference becomes useful evidence against authorship when its significance has been established through comparison. Its mere existence is not enough. The study documents changed words, altered clauses, combined material, and flexible applications in the writings normally used as the control corpus.
If a feature in an accepted work is explained by adaptation or quotation freedom, an equivalent feature in a disputed work cannot be rejected without further argument. A different judgment may be justified by the scale of the change, its function, the kind of document, or other evidence. That justification must be stated rather than supplied silently by the prior authenticity label.
Resemblance needs the same discipline. A phrase can be compatible with Clement without proving his authorship. An imitator could reproduce recognizable language. But a claim of imitation needs evidence capable of distinguishing imitation from ordinary authorial usage. Both similarity and difference must be evaluated within a comparison that allows competing conclusions.
The study supplies grounds for retesting arguments that depend on a rigid Clementine profile. It is not a statistical rerun of vocabulary-based authorship studies, and it does not invalidate every argument made from style or theology. It identifies the variation that any such argument must handle.
7. The consequence for Secret Mark is a change in the comparison standard
The strongest conclusion for the controversy is that an inflexible image of Clement is an inadequate basis for exclusion. A citation’s departure from a familiar Gospel form, a combination of scriptural language, or a changed divine designation must be compared with the behavior already present in the accepted writings.
This affects the weight that can be placed on particular objections to the Letter to Theodore. The study gives reasons to examine whether an alleged incompatibility is actually outside the observed range. It does not establish that every contested feature of the letter falls within that range. Each objection needs its own comparison.
The letter was excluded from the material used to establish the citation profile. Its authenticity was therefore not assumed in assembling the evidence. The comparison remains available for testing the disputed document rather than depending on it.
The study also leaves the letter’s provenance, date, handwriting, and historical authorship to their appropriate evidence. It does not prove that the letter is ancient, that Clement wrote it, or that its Gospel account is historically accurate. It likewise does not identify the longer Markan text with Marcion’s Gospel or another named Gospel.
What it establishes is that a claim of incompatibility must confront the variability of the accepted comparison corpus. That is a narrower result than a verdict on authenticity, but it changes an important part of the argument.
8. Clement must be reconstructed from the documents, with their differences intact
The study’s conclusions concern the relation between evidence and the historical figure reconstructed from it. “Clement” is a useful designation for organizing the material. The designation cannot guarantee in advance that every attributed passage has the same text, purpose, voice, or theological function.
A sound comparison begins with a located passage and an identified kind of statement. It distinguishes the words quoted from the explanation attached to them. It allows a continuous treatise, a polemical report, an excerpt collection, and a later fragment to contribute evidence without treating them as equivalent witnesses.
The numerical categories require the same care. The 54 groups with changed wording are evidence of differing formulations in the supplied material. Establishing the cause of each difference requires further textual work. The unresolved Greek references and broad range associations remain limits on what can be counted as a verified quotation.
The historical picture that emerges can include recurring commitments and recognizable habits. It also has to include adaptation, combination, different voices, and the uneven survival of the works. An authorial model gains explanatory value by accounting for that evidence.
The principal conclusion of The Myth of Clement is that the Clement used to judge a disputed text must first be tested against the surviving texts attributed to him. The accepted corpus cannot be allowed a range of variation that is then denied, without argument, to the disputed document. Clement must be reconstructed from the documents, with their differences intact.
Basis of this post: the verse-by-verse comparisons and source qualifications in The Myth of Clement. The findings concern the transmitted passages examined there, rather than a fresh collation of manuscripts. The source links above open the Greek texts used in the study. The earlier study includes the full comparison catalogue, agreements, unresolved cases, and bibliographical references.
