I am an atheist. I do not believe Christianity has established its supernatural claims. That does not mean I automatically applaud every argument that ends with Christianity being wrong, obsolete or irrelevant.
A conclusion does not become well supported because I happen to like where it lands.
Richard Carrier’s October 10 announcement of a debate with Jason DeCosta, set for October 18, caught my attention for precisely that reason. According to Carrier’s account, DeCosta generally interprets the New Testament’s Gentile mission as the recovery of scattered Israelites, while allowing some actual Gentile converts. He also regards the promised apocalypse and salvation as accomplished in the first century, leaving Christianity with no saving purpose today. [1]
That strikes me as three steps that each need more support than this account gives them.
A qualification belongs up front: I am responding to the position as Carrier describes it. His announcement is not DeCosta’s complete case, and Carrier is his opponent. I am not reviewing DeCosta’s own books or papers here. What I can identify are the questions this account leaves unanswered.
First, Give the Position Its Due
An Israel-centered reading does not come out of nowhere. Matthew 15:24 depicts Jesus describing his mission in terms of Israel’s lost sheep. Matthew 24:34 places the fulfillment of the events under discussion within the lifetime of a generation. [3]
Those passages deserve attention. A critic cannot dismiss them simply because they complicate a preferred interpretation.
But an initial mission to Israel does not, by itself, establish that every subsequent mission was restricted to Israelite descendants. And an expectation of fulfillment within a generation does not establish that fulfillment actually occurred.
Those distinctions are where my doubts begin.
The First Step: Who Were These Gentiles?
If DeCosta wants us to understand apparently non-Israelite audiences as scattered Israelites, he needs to show why the particular passages support that identification.
Romans 11 presents an obvious test. Paul addresses Gentiles and distinguishes them from his own people. He then describes their inclusion through the image of a wild olive shoot grafted into a cultivated tree, contrasting that process with natural branches being restored. [4]
As a lay reader, I see a distinction between outsiders being included and members being restored. DeCosta may have an explanation for why both categories belong within a broader Israelite restoration. But that explanation needs to account for the distinction, not merely repeat the theory.
Ephesians 2:11-19 raises a related question. It depicts its Gentile audience as previously excluded from Israel’s community and estranged from its covenants, then brought near. [2] Why should we understand that as the recovery of lost ancestry rather than the inclusion of outsiders?
Allowing some genuine Gentile converts is not automatically a fatal concession. A movement can have a primary audience and still accept other people. But it makes a clear method essential: which audiences were which, and what evidence tells us?
A passage-by-passage case using language, context and relevant ancient evidence. The rule for identifying Israelites must work even where its result is inconvenient for the theory. Simply saying that scattered Israelites existed does not identify the recipients of a particular letter.
The Second Step: Expected Soon Does Not Mean Happened Already
The timing language is worth taking seriously. I have no interest in making an awkward prediction disappear to protect Christianity.
But consider the choices. A writer might have expected an event that failed to happen. A passage might describe a political catastrophe through religious imagery. Different writers might have held different expectations. Or a proposed fulfillment might genuinely fit the text.
Those possibilities have to be compared. A near-term expectation cannot select its own successful outcome.
First Thessalonians 4:13-18, for example, describes the Lord’s descent, the resurrection of dead believers and living believers being gathered with them. [5] If those descriptions refer to a first-century fulfillment, what exactly occurred, and why is that what the passage means?
Symbolic language is a legitimate possibility. The question is what establishes the symbolism and its referent. An interpretation cannot become reliable merely because it makes fulfillment possible.
Define the proposed events, explain the relevant imagery through its literary and historical context and identify evidence that supports the correspondence. Where a claim concerns an invisible supernatural event, distinguish a theological interpretation from an independently demonstrated occurrence.
The Third Step: Fulfilled Does Not Mean Expired
Even if I granted the previous claims for the sake of argument, the final conclusion would still need support.
Why would completion make Christianity unavailable or meaningless to later generations?
A builder finishes a bridge. People do not stop crossing it because construction is complete. The completion is what makes its continuing use possible.
That analogy proves nothing about salvation. It simply exposes the missing step: a completed act can have continuing consequences.
If DeCosta’s argument is that salvation belonged exclusively to a particular population before a particular cutoff, then the exclusivity and the cutoff need to be established. Neither follows automatically from fulfillment.
And “offers nothing” needs precision. Having no demonstrable supernatural benefit is different from having no community, emotional, ethical or cultural value for anyone.
Evidence establishing both the limits on eligibility and the termination of access, not merely passages anticipating completion. The argument must also explain why an enduring benefit is excluded.
Three Questions That Must Stay Separate
What did the authors mean? Were the things they believed true? What follows for people now?
Those are three different questions.
Someone could persuade me that an ancient author expected salvation only for a limited group without persuading me that anyone was supernaturally saved. Someone could demonstrate that an author anticipated an imminent apocalypse without demonstrating that the apocalypse happened.
That matters especially here. A historical argument about Christianity’s original scope could be sound even if its supernatural premises were false. Conversely, rejecting those premises would not establish DeCosta’s particular account of that scope.
My Bottom Line
As described in Carrier’s announcement, DeCosta’s position leaves me unconvinced. It needs firmer support for the identity of its audiences, the alleged fulfillment of its promises and the conclusion that those promises have no continuing application.
I would reconsider if the evidence justified those steps. That is what skepticism requires.
I do not need DeCosta to be right for my atheism to survive. I also do not need him to be wrong to protect Christianity. I want the argument to survive examination.
Why This Matters
Skepticism becomes selective when we demand evidence from religious believers and accept convenient conclusions from their critics.
If I criticize someone for assuming the answer before examining the evidence, I cannot excuse the same method because the resulting argument happens to undermine Christianity.
My standard is simple: explain the text, support the historical claims and show how the conclusion follows.
I am an atheist. I still expect an argument to earn its ending.
- Carrier, Richard. (2026, October 10). “Debating Israel Only with Jason DeCosta.” Richard Carrier Blogs. richardcarrier.info/archives/44691. Used as the description of the position under discussion, not as a substitute for DeCosta’s complete argument.
- Ephesians 2:11-19. New Revised Standard Version Updated Edition.
- Matthew 15:24 and Matthew 24:34. New Revised Standard Version Updated Edition.
- Romans 11:13-24. New Revised Standard Version Updated Edition.
- 1 Thessalonians 4:13-18. New Revised Standard Version Updated Edition.










