Luke at Padua: What the Bones and DNA Can Tell Us

Jerome's On Illustrious Men, chapter 7, connects Luke with Antioch and Paul and reports a transfer of his bones to Constantinople under Constantius. We consulted the published English translation, not a manuscript or a new critical edition. It is evidence that this account existed in late antiquity; it is not an eyewitness record of Luke's death or a document tracing the skeleton now in Padua. [SRC-04]

In Against Vigilantius, section 5, Jerome also invokes the transfer of Andrew, Luke and Timothy while defending the honoring of relics. The passage belongs to a polemical argument. It shows that the public remembrance of such transfers could serve a theological discussion, but it does not give an inventory that can be compared bone by bone with a modern examination. The work's aggressive descriptions of an opponent are not adopted here. [SRC-05]

A historian must therefore distinguish the existence of a tradition from the continuity of the object to which a later community applies it. A report naming a saint in one city supplies a starting point. To reach another city centuries later, one needs further documents, dates, descriptions and reasons for identifying the objects in those documents as the same remains. Drawing a line between the cities cannot supply a missing record.

This does not make ancient testimony worthless. It gives a concrete question to investigate: which parts of the later account can be connected to an earlier source, and where does the connection become an inference? The best reading preserves both the testimony and the gap. It neither treats every old narrative as a shipping receipt nor assumes that an incomplete record must have been invented.

Santa Giustina dates the stone sarcophagus to 1313 and its present placement to 1562. Its account offers alternative periods for the earlier arrival at Padua. The modern examination began with an opening on 17 September 1998; results were presented at the 2000 congress. These are different dates for different events. [SRC-02]

A visitor can easily collapse them into one story: an old tomb must contain an equally old burial, and an equally old burial must belong to the person named on the shrine. Neither inference follows automatically. The age of a monument concerns its construction. The age of a skeleton concerns a biological history. An inscription concerns an attribution made or preserved by someone. Connecting all three requires evidence in addition to their present proximity.

The distinction also protects the value of the shrine. Its architecture, monastic custody and devotional use remain historical subjects even if personal identification is unresolved. A church is not reduced to a mislabeled laboratory sample, and a sample does not acquire a biography from the beauty of its setting. Keeping the questions separate allows the material and religious histories to be studied more carefully.

The full medieval custody dossier and the original records of every opening were not inspected for this article. A useful next stage would compare inventories across openings, recording missing elements and changes of container without silently harmonizing them. Such a comparison could strengthen or weaken particular links. It would still need to explain how those links reach the named individual, rather than merely an earlier attribution.

The 2001 PNAS paper describes an elderly male and reports an AD 72–416 radiocarbon interval at 95% confidence in Figure 2, citing personal communications. Individual laboratory reports and calibration details are not reproduced. [SRC-01]

Age at death and calendar age should never be merged. The first concerns how long a person lived. The second concerns when the sampled material formed. An assessment that someone was elderly does not locate that life in the apostolic period; an interval overlapping that period does not establish an exact year of death. An overlapping interval can leave many historical candidates possible.

The published date is relevant evidence, but the form of its reporting limits independent checking. Readers can verify what the authors printed. They cannot use that caption alone to reconstruct each laboratory's measurement, sample preparation or calibration. We therefore retain the interval as a reported result and do not offer a recalculated date, a new confidence estimate or an unsupported claim that every bone was dated.

There is an important asymmetry here. If a securely tested specimen falls wholly outside a person's lifetime, the result can exclude that person's identity for that specimen. Falling within a broad compatible period is less decisive: many people shared it. Compatibility removes one obstacle; it does not automatically overcome the remaining questions of custody, sampling and personal identification.

Only one of two teeth yielded sufficient DNA. The authors assembled 360 mitochondrial bases using overlapping fragments and 39 clones. Independent extraction replication was unavailable because material was scarce, despite internal checks. [SRC-01]

Those distinctions are easy to lose in a headline. Sampling two teeth is not the same as obtaining two independent successful profiles. Making multiple copies within an experiment is not equivalent to repeating extraction from separate material in another laboratory. Internal consistency and independent replication address related questions, but they do not have identical evidential force.

A sequence is also a description of genetic material, not a personal name. Imagine finding a short passage in an unsigned letter. Its language might be informative about a broad community, but knowing the language would not identify its author. The analogy is limited, yet it helps locate the missing step: a group resemblance must not be promoted into an individual identification without additional evidence connecting that individual to the sample.

The original paper is available in full and was read with its tables and figures. That access does not mean we handled the tooth, repeated the experiment or inspected the original laboratory notebooks. The article's judgment concerns the published evidence and its limitations. New testing would require its own permissions, preservation assessment and scientific plan; it cannot be promised as an automatic solution to an old attribution.

Reference samples comprised 49 modern Syrians, 48 Greeks and 96 Anatolians. The reported Syria/Greece odds were 2.87, with an empirical 95% interval of 1.06–6.17; Syria/Anatolia odds were 1.41, with 0.66–2.88. Table 1 counts pairwise sequence comparisons. [SRC-01]

The practical question is what belongs on either side of those ratios. Luke and everyone else are not the two alternatives in the calculation. Modern reference groups are. The output cannot be converted into a percentage chance that the skeleton is the evangelist. Doing so changes the question after the calculation has been performed, while retaining the appearance of mathematical precision.

An invented example makes the problem clearer. Suppose an unsigned note resembles writing in collection A more than writing in collection B. That comparison could guide further investigation. It would not establish which writer in A produced the note, whether an unexamined collection C contains closer examples, or whether the note really belongs to the historical period under investigation. A favorable comparison can be informative without being individually identifying.

The reference populations are modern, whereas the attribution concerns antiquity. The relationship between them must be argued, not assumed from shared place names. Likewise, the many pairwise comparisons in a table do not create that many ancient people or independent relic samples. A large denominator can come from repeatedly comparing members of a modest reference collection. Reading the table heading is therefore part of understanding the result.

The original analysis leaves substitution in Constantinople compatible with its findings. [SRC-01] That surviving alternative matters: a result offered in support of a traditional narrative can still fail to distinguish the narrative from an important rival. The responsible conclusion must retain the alternative instead of quoting only the comparison that sounds most favorable.

Bandelt and colleagues' 2006 chapter, pages 131–132, questions both the geographical inference and the reported combination of control-region changes with the AluI restriction result. They note that the surrounding restriction-site region was not cloned and raise the possibility that a minority contaminating component affected the band. This is a methodological criticism of the published work, not a new analysis of the Padua remains. [SRC-06]

That criticism prevents an overly simple account in which the only uncertainty is whether an unquestionably authentic ancient sequence has enough geographical resolution. Authenticity of the recovered sequence is itself a question. The argument deserves consideration even though it does not supply an alternative sequence, an independently tested skeleton or the name of another person.

We should also avoid reversing the overclaim. A published concern is not permission to announce that a particular contamination event has been demonstrated in a particular laboratory. The authors of the criticism explain a vulnerability and question the inference. Determining what actually occurred would require further evidence. The distinction between identifying a possible failure mode and proving that it occurred is essential to fair scientific reporting.

Our assessment therefore does not treat the original paper as a certificate or the later chapter as a verdict on the researchers' intentions. It asks whether the available evidence closes the identification problem. It does not. A full specialist review would examine the original analytical data and subsequent scholarship, including any direct responses, rather than turning disagreement into a contest between devotional and skeptical personalities.

John Paul II's message of 17 October 2000 describes Padua's custody according to tradition. It links the congress with an ecumenical request from the Orthodox Archbishop of Thebes and directs attention to Luke's Gospel and Acts. The Diocese of Padua's 2025 anniversary notice still speaks of the body attributed to Luke. These are records of ecclesial remembrance and its stated scope. [SRC-03; SRC-07]

The careful wording should survive translation into a public article. A papal message welcoming an investigation does not certify every technique, close every historical gap or assign a genetic probability. Nor does the absence of such a certification imply that the Church has withdrawn the saint from veneration. The subject of the devotion and the historical identification of a particular set of remains are distinguishable.

Jerome's defense of relics rejects treating a human being as God. [SRC-05] For Catholic readers, that distinction remains a helpful guard against thinking of material remains as possessing independent divine power. Prayer seeks communion with God; honoring a saint remembers a life received from God. Neither purpose requires us to describe a scientific uncertainty as settled.

A person visiting Santa Giustina can attend to the Gospel, the community praying there and the history of the shrine while acknowledging the limits of the identification. A person reading from outside Catholic faith can examine those same records without being asked to accept devotion as laboratory evidence. Clarity makes room for both encounters, because it allows each claim to be considered on its own terms.

The editorial verdict is UNRESOLVED for the personal identification of the Padua individual as Luke the Evangelist. This does not mean every aspect of the story is equally uncertain. There are accessible ancient texts, a present custodian, a documented modern investigation and a published scientific debate. What remains unestablished is the connection that turns this individual into that named person.

Further work should be aimed at that connection. Historical research could clarify the earliest Padua attribution and changes in custody. Scientific review could revisit the original data, sample history and feasibility of independent work. Each proposal should explain which competing possibility it could actually distinguish. More measurements are not automatically more identification if they continue to answer only broad questions about age or affinity.

The Catechism's respect for the dead is not restricted to remains carrying a famous name. [SRC-08] That has a practical implication for this case: uncertainty does not make the individual disposable. Preservation, investigation and public explanation should remember that the subject was a person, even where the accessible record cannot recover that person's name.

Luke's prologue presents an author concerned with received testimony and an orderly account. Readers may return to Luke 1:1–4 without treating it as a description of modern laboratory practice. [SRC-09] In our Catholic reading, attention to what a witness says, what has been handed on and what remains to be established is an appropriate response to this case. Honest limits can deepen attention rather than end it.