James 2:19 is the verse people reach for when they want to say that belief is not enough. “You believe that God is one; you do well. Even the demons believe—and shudder!” The inference is supposed to be obvious: demons believe and are damned, therefore belief of that kind does not save, therefore something must be added—repentance, submission, obedience, perseverance—before faith becomes the sort of faith that justifies.
However, the inference is not obvious. It is not even valid. And when the passage is read on its own terms, it turns out to be arguing something rather different, and in places the opposite.
What follows is not a defence of a slogan. It is an attempt to read James 2:14–26 as a coherent argument written by a Jewish Christian to Jewish Christians about a problem in their assemblies, and to see what the demons are actually doing in it.
The argument that has to be answered
It is worth stating the case against faith alone in its strongest form, because the weak form is easy to knock down and knocking it down proves nothing.
The weak form says: the demons believe the wrong thing—monotheism rather than the gospel—so their case is irrelevant. This is true as far as it goes, but a competent opponent will not be detained by it, because his point was never about the content.
The strong form says: James chose the most minimal, most orthodox, most unimpeachable proposition available—the Shema itself—precisely in order to isolate the mode of believing from its content. His argument is that bare intellectual assent, however correct its object, is not what Scripture means by saving faith. Assent is the genus that demons and unregenerate professors share. Saving faith is assent plus trust plus surrender, and the works of 2:14–26 are the necessary and inevitable evidence that the further elements are present. Whatever proposition you plug in, the manner of holding it is what James is testing.
That argument has to be answered on its own ground. Three things answer it: James’s own vocabulary, James’s own grammar, and James’s own simile.
What “save” means in James
Everything in this passage turns on 2:14: “Can that faith save him?” If σῴζω here means “justify before God and grant eternal life,” the passage is at least prima facie about the way of salvation. If it does not, the entire debate has been conducted on a false premise.
James uses σῴζω five times, and he is the best commentator on his own usage.
1:21 — “receive with meekness the implanted word, which is able to save your souls.” Three verses earlier he has said that God “brought us forth by the word of truth” (1:18). His readers have already been born of the word. Whatever the word is now able to save them from, it is not the state of being unregenerate. They are being told to receive something they already possess in principle, and the saving lies ahead of them.
2:14 — the disputed instance.
4:12 — “There is one lawgiver and judge, he who is able to save and to destroy.” Note the phrasing: one lawgiver, εἷς ἐστιν, the same construction as the confession in 2:19. James’s monotheism is not a doctrine he holds at arm’s length; it is the ground of his warning about judgement.
5:15 — “the prayer of faith will save the one who is sick.” Physical deliverance. Nobody reads this as justification.
5:19–20 — “if anyone among you wanders from the truth and someone brings him back, let him know that whoever brings back a sinner from his wandering will save his soul from death and cover a multitude of sins.” A brother. Within the community. Wandering. And the person restoring him “saves a soul from death”—the same collocation as 1:15, where sin, fully grown, gives birth to death.
Four of James’s five uses are unambiguously about deliverance from consequences rather than transfer from condemnation to justification. The fifth is the one under dispute. On any normal principle of reading an author by his own idiom, 2:14 belongs with the other four.
The immediate context settles it further. James 2:14 does not begin a new subject; it follows directly from 2:12–13: “So speak and so act as those who are to be judged under the law of liberty. For judgement is without mercy to one who has shown no mercy.” That is a warning to believers about their assessment, and it is the sentence to which “what use is it, my brothers” is the immediate sequel. The illustration that follows in 2:15–16—a brother or sister sent away hungry with a blessing—is mercilessness in exactly the sense of 2:13.
James is not asking whether a professing Christian is really a Christian. He is asking what use a believer’s faith is to him when he stands before the Judge having done nothing with it.
The demons: what the comparison actually says
Now the verse itself.
σὺ πιστεύεις ὅτι εἷς ἐστιν ὁ θεός; καλῶς ποιεῖς· καὶ τὰ δαιμόνια πιστεύουσιν καὶ φρίσσουσιν.
Three features of the sentence are routinely passed over.
James commends the belief. καλῶς ποιεῖς—”you do well.” There is an edge to it, certainly, but James does not say the man is deceived, does not say his belief is counterfeit, does not say he only claims to believe this. He affirms it and moves on. Whatever is wrong here, it is not that the proposition is false or that the man’s grasp of it is fictitious.
The demons are the greater, not the lesser, case. καὶ τὰ δαιμόνια πιστεύουσιν καὶ φρίσσουσιν—they believe and they shudder. φρίσσω is a New Testament hapax denoting a shuddering, bristling dread. The rhetorical force of the second καί is ascensive: the demons’ belief at least produces a reaction. The interlocutor’s produces nothing at all. He is being told that his orthodoxy is more inert than a demon’s.
This is mockery, not taxonomy. James is not classifying species of faith into saving and non-saving; he is ridiculing a man whose confession achieves less than that of beings who are beyond hope. Reading the verse as a doctrinal definition of “demon faith” requires ignoring what the sentence is doing.
The verse says nothing about how the demons are lost. It says they believe and shudder. It does not say they are lost because of the character of their believing. That has to be imported.
Why the demons prove nothing about faith
Here is the syllogism the passage is usually asked to support:
- The demons believe.
- The demons are not saved.
- Therefore belief of this kind cannot save.
Step 3 does not follow, because the demons’ condition is independently and sufficiently explained by something that has nothing to do with the quality of their assent. There is no atonement for angelic beings. Hebrews 2 is clear. The Incarnation is bounded to one order of being, and it is bounded deliberately—the argument of Hebrews 2:14–17 is that the Son took flesh and blood because the people he came to help have flesh and blood. Fallen angels are held for judgement (2 Peter 2:4; Jude 6); no substitute has died in their place; no offer has been extended to them. There is no salvation available for demons.
This does not tell us what the passage means. It is a defeater for an inference, and nothing more. But it is a decisive defeater. You cannot reason from the non-salvation of a class of beings for whom salvation was never provided to a conclusion about the sufficiency of faith for beings for whom it was.
There is a sharper corollary. If saving faith is trust in Christ’s death on one’s own behalf, then the demons could not exercise it under any conditions whatever—not through defective sincerity, but because the proposition is not available to them. “Christ died for me” is not a thing a fallen angel can believe, truly or falsely. The demons are therefore useless as a test case for what human faith must contain, since the one thing human faith rests on is a thing they cannot be asked to rest on.
The mode objection: assent and trust
Which brings us to the strong form of the objection. Grant everything above: the opponent still says that James picked the Shema precisely so that content would drop out and only the manner of believing would remain in view. Bare assent, he says, is what demons do. Saving faith is something else in kind.
The difficulty is that this distinction is a systematic construct, not a lexical one. The tripartite scheme—notitia, assensus, fiducia—is Protestant scholastic taxonomy, useful for catechesis and entirely absent from the vocabulary of the New Testament. Greek does not have one verb for assenting and another for trusting. It has πιστεύω, and πιστεύω does both jobs.
More damagingly for the objection: the specific construction it treats as the marker of sub-saving belief—πιστεύω followed by ὅτι and a proposition—is the construction the New Testament uses for saving faith.
- John 20:31 — “these are written that you may believe that Jesus is the Christ, the Son of God, and that believing you may have life in his name.”
- John 11:27 — Martha: “I have believed that you are the Christ, the Son of God.”
- 1 John 5:1 — “everyone who believes that Jesus is the Christ has been born of God.”
If πιστεύω ὅτι denoted a mode of believing too shallow to save, John’s stated purpose in writing his Gospel would be incoherent and 1 John 5:1 would be false. The construction is not the problem.
So the difference between the demons and the believer cannot be located in the manner of believing. It has to be located in the object. And that is where it in fact lies: the Shema is not the gospel. To affirm that God is one is to affirm something demons affirm and something Pharisees affirmed; it is not to trust in a crucified and risen Christ for the forgiveness of sins. The object argument, which is weak when asserted at the outset, is strong when arrived at by elimination.
Dead faith is real faith
The decisive textual argument is the one the passage ends with, and it is usually the one left out.
ὥσπερ γὰρ τὸ σῶμα χωρὶς πνεύματος νεκρόν ἐστιν, οὕτως καὶ ἡ πίστις χωρὶς ἔργων νεκρά ἐστιν. (2:26)
“For as the body apart from the spirit is dead, so also faith apart from works is dead.”
A corpse is a body. That is the whole force of the comparison. James’s chosen analogue for workless faith is not a mirage, not a forgery, not an empty claim, but a real thing that has ceased to function. Whatever else νεκρός means in 2:17, 2:20 and 2:26, it cannot mean “non-existent,” because the term of comparison James supplies to define it is something indisputably existent and indisputably useless.
The textual tradition reinforces this. In 2:20 the reading of Vaticanus, the original hand of Ephraemi, and 1739 is not νεκρά but ἀργή—”idle,” “unemployed,” “unproductive”—and it is almost certainly original, νεκρά being an assimilation to the surrounding verses (see Metzger’s Textual Commentary, ad loc.). ἀργή is a compound of ἀ- and ἔργον: James is punning. Faith without works is un-working. The word is used of capital that yields no return and of land left fallow. Money in a mattress is money. Fallow ground is ground.
The point is not marginal. It is the difference between two readings of the entire passage:
- Faith without works is not faith. This makes works constitutive or evidentially necessary, and it is what the passage is usually taken to teach.
- Faith without works is faith that accomplishes nothing. This is what James’s own simile and his own word choice say.
It is worth noting that the strongest exponents of the opposing view concede more here than they may realise. Moo states the contrast as one between a faith that has works and a faith that does not, the latter being lifeless like a body without a spirit, and profiting nothing at the day of judgement (James, TNTC, 104). That is very nearly the argument being made here. “Profits nothing at the day of judgement” is precisely the position: the faith is real, and it will not deliver its holder from anything at the assessment. The only step remaining is to stop equating “profits nothing at judgement” with “is not saving faith”—a step James never takes, and which requires the assumption that the only thing at stake at the judgement of believers is entrance.
James 2:24
“You see that a person is justified by works and not by faith alone.” It is the only occurrence of “faith alone” in Scripture and it is negated, and any treatment of this passage that avoids the verse deserves to be disbelieved.
Two observations.
First, James never says “justified before God.” Paul does, repeatedly and pointedly (Romans 3:20; 4:2; Galatians 3:11). James’s frame from 2:18 onward is demonstrative: “show me your faith apart from your works, and I will show you my faith by my works.” The verb δικαιόω has an established range that includes vindication and being shown to be in the right—it is used this way of God himself (Luke 7:29; 1 Timothy 3:16) and of Wisdom (Matthew 11:19). To be justified in James’s sense is to have one’s faith made visible and acknowledged.
Second, James tells us the relationship between the two justifications in his own words. Genesis 15:6 was, he says, fulfilled at Moriah: καὶ ἐπληρώθη ἡ γραφή (2:23). A text is fulfilled when what it declared comes to its full expression. The declaration came first; the act brought it to completion. James is not saying Abraham became righteous at Moriah. He is saying Moriah is what the Genesis 15 verdict looked like when it finally worked itself out.
The most serious opposing reading takes James’s δικαιόω as eschatological vindication at the final judgement rather than vindication before human observers. That reading has real strength, and it fits the judgement framing of 2:12–13 better than the “before men” reading does. But it does not help the case against faith alone, because it still has to reckon with ἐπληρώθη: whatever happens at Moriah is the outworking of a verdict already rendered, not the securing of one.
Abraham and Rahab
James’s two examples are frequently deployed as though they were parallel demonstrations that faith and works arrive together. Read against their narratives, they are demonstrations of the opposite.
Abraham. The gap between Genesis 15:6 and Genesis 22 is very large. Genesis 15 precedes Genesis 16, and Abram is eighty-six at Ishmael’s birth (Genesis 16:16), so the reckoning of righteousness falls no later than his eighty-fifth year. Isaac is born when Abraham is a hundred (Genesis 21:5). At Moriah, Isaac carries the wood for the burnt offering up the mountain and conducts a conversation with his father (Genesis 22:6–7)—not a small child. Josephus puts him at twenty-five (Antiquities 1.227); rabbinic tradition, reasoning from Sarah’s death at 127 shortly afterwards, puts him at thirty-seven. Either way the minimum interval is around three decades and the likely figure is closer to forty years.
Forty years of justified standing before the work that “justified” him. If those are the same justification, Abraham spent four decades in an indeterminate state that Genesis 15:6 declares to be settled.
And what the obedience at Moriah secured is stated in the text: an oath-confirmed reaffirmation of blessing, “because you have done this and have not withheld your son” (Genesis 22:16–18), and, in James’s own phrase, that he “was called a friend of God” (2:23). Friendship, and reward. Not entry.
Rahab. Joshua 2 gives the sequence plainly, and it is not the sequence usually claimed for it. She takes the men and hides them in 2:4, and conceals them under the flax on the roof in 2:6. Only after the pursuers have gone out does she go up to them and make her confession (2:8–11). The act precedes the speech.
This does not damage the argument; it strengthens it, once the confession is read for what it says. “I know that the LORD has given you the land… we have heard how the LORD dried up the water of the Red Sea before you… and as soon as we heard it, our hearts melted” (Joshua 2:9–11). The drying of the sea was forty years earlier. Sihon and Og were recent. Rahab is not reporting a conviction formed in the last ten minutes; she is reporting a settled belief formed from reports long before the spies knocked. Her concealment of them was already an expression of that belief. The speech articulates what the act had assumed.
Hebrews 11:31 makes the ordering explicit: “By faith Rahab the prostitute did not perish with those who were disobedient, because she had given a friendly welcome to the spies.” Faith is the ground; the welcome is what faith did.
And note what Rahab’s works accomplished, because it is exactly what James means by σῴζω. They did not obtain her eternal life. They got her and her household out of Jericho alive (Joshua 6:22–25). James’s second worked example is a case of deliverance from temporal destruction through action taken in faith. That is the whole thesis of the passage, illustrated.
What this does not mean
None of this makes works optional, and the argument is not improved by pretending otherwise.
James thinks a believer whose faith produces nothing is in serious trouble. He says so with some violence: such a man is ἀργός, useless; he is κενός, empty-headed (2:20); he faces judgement without mercy (2:13); his sin, matured, gives birth to death (1:15); he may need another believer to drag him back from it (5:19–20). A faith that does nothing is a scandal and a danger, and James wants it treated as one.
What he does not do is question whether the man is a believer, or make his works the ground or the proof of his standing before God. He assumes the faith is real and tells the man that it is doing him no good—which is a coherent thing to say only if faith and its usefulness are separable. If they were not, the letter would have no argument.
Summary
| Paul (Romans 3–4, Galatians 3) | James 2:14–26 | |
|---|---|---|
| Question | How is a person declared righteous before God? | What use is a believer’s faith if it does nothing? |
| Addressees | Those seeking justification through law-works | Believing brothers already born of the word (1:18) |
| Sense of δικαιόω | Forensic: declared righteous | Vindicated: shown to be what it claims |
| Sense of σῴζω | Deliverance from condemnation | Deliverance from consequence and judgement (cf. 1:21; 5:15, 20) |
| Abraham | Genesis 15:6, before circumcision, before any act | Genesis 22, three to four decades later |
| Rahab | — | Delivered from Jericho’s destruction by acting on prior belief |
| Faith without works | Justifies | Real, but idle, and profits nothing at judgement |
The demons in James 2:19 are not a warning about faith with insufficient ingredients. They are a rhetorical humiliation: your confession does less for you than theirs does for them, and theirs does nothing at all, for reasons that have nothing to do with faith and everything to do with the fact that no one died for them.
Belief in Christ for eternal life is sufficient because of what Christ did, not because of what belief is like. James never says otherwise. He says something more uncomfortable and more useful: that a believer can hold the truth and do nothing with it, and that this will cost him.
Sources and further reading
Opposing positions, in their strongest form. Douglas J. Moo, The Letter of James, Pillar New Testament Commentary (Eerdmans, 2000), and the shorter James, Tyndale New Testament Commentaries (IVP, rev. ed. 2015); Scot McKnight, The Letter of James, NICNT (Eerdmans, 2011); Craig L. Blomberg and Mariam J. Kamell, James, ZECNT (Zondervan, 2008); Dale C. Allison Jr., A Critical and Exegetical Commentary on the Epistle of James, ICC (Bloomsbury, 2013).
Genre and setting. Richard Bauckham, James: Wisdom of James, Disciple of Jesus the Sage (Routledge, 1999).
The reading argued here. Zane C. Hodges, Dead Faith: What Is It? (Redención Viva, 1987) and The Epistle of James: Proven Character Through Testing (GES, 1994); Joseph Dillow, The Reign of the Servant Kings (Schoettle, 1992), and Final Destiny (Grace Theology Press, 2012).
Text. Bruce M. Metzger, A Textual Commentary on the Greek New Testament, 2nd ed. (UBS, 1994), on James 2:20.
Leave a Reply