Sam Harris has spent more than twenty years arguing that Islamic extremists often understand Islam more accurately than Muslims who interpret their tradition in less literalistic ways.

The argument depends heavily on the Qur’an and the hadith. Harris reads passages concerning jihad, apostasy, unbelief, martyrdom and warfare, compares them with the rhetoric of groups such as ISIS and al-Qaeda, and concludes that the extremists have a disturbingly plausible case.

There is a problem underneath that argument that receives far less attention.

How did Harris determine what the hadith mean?

Not whether he read them. Not whether the quotations are real. Not whether Sahih al-Bukhari or Sahih Muslim contains disturbing material.

The question is methodological.

A hadith collection is not a catechism. It is not a systematic theology. It is not a law code in which every individual entry automatically functions as a universal command. Hadith literature consists of thousands of reports transmitted through different chains, preserved in different collections, sometimes with different wording, and interpreted through an enormous scholarly apparatus that developed over centuries.

Harris largely bypasses that apparatus while nevertheless reaching remarkably confident conclusions about what Islam teaches.

That is the problem.

What Harris Actually Did

We do not have to speculate about at least part of Harris’s method because he tells us.

In The End of Faith, Harris introduces a series of hadith about jihad and martyrdom to demonstrate what he regards as Islam’s intrinsic militancy. Among them are reports concerning the superiority of fighting in God’s cause, the reward of guarding the frontier, the martyr wishing to return to life so that he can be killed again, and the famous statement that paradise lies under the shadow of swords. After presenting this collection of quotations, Harris concludes that Islamists regularly invoke such material to justify violence.

Then comes the revealing footnote.

Harris explains that some of these hadith were taken from Bernard Lewis’s The Crisis of Islam. The others, he says, were “drawn from an Internet database”, specifically the searchable hadith database maintained at the time by the Muslim Students Association at the University of Southern California.

There is nothing inherently wrong with using a searchable hadith database. Muslims use electronic databases constantly today. Sunnah.com performs much the same reference function for millions of people.

But finding a hadith is not the same thing as interpreting a hadith.

Harris gathered a cluster of translated narrations concerning warfare and martyrdom and treated their apparent English meaning as evidence for what Islam itself teaches. What is conspicuously absent from the argument is sustained engagement with the interpretive disciplines Muslims traditionally use to determine what a hadith means, how it applies, and how it relates to other evidence.

That distinction is not minor. It is the whole issue.

A Sahih Hadith Is Not a Self-Interpreting Proposition

Hadith scholarship is much more complicated than asking whether a quotation appears in Sahih al-Bukhari or Sahih Muslim.

At the most basic level, a hadith contains an isnād, the chain through which the report was transmitted, and a matn, the content of the report itself. Scholars developed extensive disciplines for evaluating transmitters, comparing chains, examining textual variants and determining the reliability of reports.

But even establishing that a report is sahih, or sound, does not settle what the report means or how it functions in Islamic law.

Authenticity and interpretation are different questions.

A jurist may have to determine whether a report is general or limited to a particular circumstance. A seemingly absolute statement may be qualified by another narration. Two authentic reports may appear to conflict and require reconciliation. The practice of the Companions may affect how a report is understood. Qur’anic principles have to be considered. Questions of chronology may arise. The Arabic itself may permit meanings that disappear in translation.

This is where disciplines such as usul al-fiqh, the principles of jurisprudence, become relevant. Muslim jurists developed technical categories for determining how textual evidence operates. General and particular language, unrestricted and qualified statements, commands and recommendations, apparent conflicts between evidence, contextual application, consensus, analogy and other questions are part of the interpretive process.

The hadith commentarial tradition developed alongside these disciplines.

Al-Nawawi’s commentary on Sahih Muslim remains one of the major works through which that collection has traditionally been understood. Ibn Hajar al-Asqalani’s Fath al-Bari became the most celebrated commentary on Sahih al-Bukhari. These works examine language, chains of transmission, variant reports, legal disagreements, theological questions, chapter headings and apparently contradictory narrations.

This was not some marginal activity invented by modern Muslims trying to soften embarrassing passages. Hadith commentary became a major Islamic scholarly genre. Oxford’s overview of the field describes commentaries being delivered as live lessons, oral glosses and massive written works used for devotional study, legal instruction and legal practice. Commentators analyzed not only the content of reports but their isnāds and even the significance of the chapter headings under which compilers placed them.

Some forms of hadith commentary go back to the eighth century, when scholars were already producing works explaining unusual vocabulary and obscure expressions using philology, grammar and lexicography.

The idea that the correct meaning of a hadith is simply whatever appears most obvious to an intelligent person reading it in English would have been foreign to that entire enterprise.

“But Muslims Read Sunnah.com Too”

Of course they do.

That objection misses the point because the difference is not the website. It is the a priori interpretive framework brought to the website.

A Muslim browsing Sunnah.com does not ordinarily encounter the hadith tradition as a blank slate.

Long before opening the database, the Muslim has already encountered Islam through the Qur’an, prayer, khutbahs, sīra, teachers, family, religious instruction, devotional practices and some degree of exposure to fiqh. He already knows that Bukhari and Muslim are collections of transmitted reports rather than books personally written as systematic theology by Muhammad. He generally knows that scholars interpret hadith and that simply finding one narration does not settle every theological or legal question.

That Muslim may possess only a crude understanding of all this. He may interpret a hadith badly. He may even become a literalist.

But he is not normally constructing Islam from zero by searching an English database for words such as jihad, martyr, apostate and sword and then deciding what the religion fundamentally teaches.

His prior framework matters.

This is true of every mature religious tradition. A Catholic reading the Gospel of John already has ideas about Trinity, Incarnation, Eucharist, Church and tradition before opening the text. An Orthodox Christian reading Paul has inherited categories about salvation, theosis, liturgy and the Fathers. A rabbi does not approach a translated verse from Deuteronomy as though the intervening history of Jewish interpretation never happened.

None of this means inherited interpretations are automatically correct.

It means texts are never encountered without interpretation.

Harris Thinks Literalism Is the Neutral Reading

This is where Harris’s deeper hermeneutical assumption becomes visible.

In a published conversation with journalist Graeme Wood, Harris explicitly argued that while scriptures permit multiple interpretations, the more “plausible,” “straightforward” and comprehensive readings tend to be “the more literalistic.” He then said that, to his eye, ISIS was offering a very plausible reading of the Qur’an and hadith.

That tells us exactly what Harris is doing.

Literalism is not merely one interpretation he is evaluating. It is being granted an epistemological advantage because Harris regards it as closer to the obvious meaning of the words.

Traditional interpretation then becomes suspicious precisely because it is interpretation.

Contextualization looks like evasion. Legal qualification looks like softening. Mystical interpretation looks like moving away from the text. Centuries of jurisprudence can begin to look like Muslims constructing elaborate excuses not to believe what their religion plainly says.

Meanwhile, Harris’s own reading disappears from view.

It becomes simply “what the text says.”

But Harris has a hermeneutic too.

He has assumptions about what counts as plain meaning, how divine commands should function, whether historical context limits a statement, how apparently conflicting texts should be handled, which reports deserve emphasis and what constitutes an honest reading.

Removing the Muslim interpretive tradition does not remove interpretation.

It merely removes Muslim interpretation.

What remains is Sam Harris interpreting an English translation.

Harris Still Makes the Same Argument

This is not merely a criticism of a book Harris wrote in 2004.

In a 2024 EconTalk interview, Harris openly acknowledged that he is not an Arabic speaker and cannot read the Qur’an in Arabic. He then dismissed the idea that this prevents a person from grasping Islam’s essential message. His advice to listeners was essentially to read a plausible English translation themselves. He argued that the Qur’an is short enough to read in a weekend and that readers will absorb its central message without knowing Arabic.

That position is remarkably consistent with the method visible in The End of Faith.

Read the translation. Observe what appears obvious. Distrust interpretations that move too far from that obvious meaning.

Harris then uses the result to evaluate the religious understanding of nearly two billion Muslims.

The confidence involved is difficult to justify.

Someone could read the Bible in English over a weekend as well. He could search for slay, wrath, stone, sword, hell and unbeliever. Every quotation would be genuine. Nothing would have been fabricated.

But the resulting document would not constitute a serious account of Christianity.

The problem would not be the quotations.

The problem would be the method.

The Fundamentalist Feedback Loop

Harris’s method also creates a curious feedback loop.

He begins with the assumption that the most literal reading is generally the most straightforward. He then reads Islamic texts with that preference for literalism. Unsurprisingly, the interpretations produced by fundamentalists often resemble the interpretations his method produces.

Their agreement then becomes evidence that the fundamentalists are reading Islam honestly.

The reasoning is circular.

ISIS reads a text literally.

Harris believes literal readings are usually more straightforward.

Harris therefore finds the ISIS reading plausible.

The plausibility of the ISIS reading then becomes evidence that Islam itself contains the problem Harris identified at the beginning.

What has actually been demonstrated is that two literalist hermeneutics sometimes generate similar readings.

It does not follow that either has exhausted the meaning of the Islamic tradition.

There is an irony here. Harris has frequently criticized religious people for bringing assumptions to texts and then finding those assumptions confirmed by the texts.

Yet his approach to hadith does something remarkably similar.

He approaches Islamic sources believing literalism reveals religion in its least disguised form. He selects passages whose literal readings support his theory. He largely bypasses the interpretive machinery that produces alternative readings. Then he points to the result as confirmation that literalism was the correct interpretive key.

What Harris Actually Proved

None of this requires pretending difficult hadith do not exist.

They do.

Islam has texts concerning warfare, punishment, apostasy, political authority and relations with non-Muslims that deserve serious examination. Muslim history contains coercion, violence, empire and theological arguments that cannot simply be wished away.

The argument is not that every uncomfortable hadith becomes harmless once a scholar explains it.

The argument is much narrower and much stronger.

You cannot determine what Islam teaches merely by locating authentic quotations and deciding what their English translations appear to mean in isolation.

Islamic civilization itself never treated hadith that way.

It created hadith criticism, jurisprudence, legal theory, philology, commentaries, theological schools and competing interpretive traditions precisely because transmitting the words was only the beginning of the problem.

Sam Harris certainly read hadith.

The more important question is whether he learned how hadith are read.

Based on the methodology he documented in The End of Faith, and the theory of interpretation he has continued to defend since then, there is good reason to think he confused access to the texts with mastery of their meaning.

And from that confusion came one of the central claims of his critique of Islam: that fundamentalists are often the Muslims who have understood their religion most honestly.

What Harris actually demonstrated is something much more limited.

When Sam Harris reads selected Islamic texts in English, privileges literal interpretation, and removes most of the traditional machinery Muslims use to interpret those texts, Sam Harris arrives at something that looks a great deal like fundamentalism.

That is an interesting fact about his hermeneutic.

It is not the same thing as proving that his hermeneutic is Islam.