Key Takeaways
Defend belief with reasons, but make gentleness part of the argument.
A defense, not an apology. Craig defines apologetics, from the Greek apologia, as a reasoned defense of Christian belief. The aim is not to quarrel, but to explain why a claim is credible. A logical argument is a sequence of premises leading to a conclusion. Its structure can be valid even when a premise is false, so a good defense must support its premises as well as show how they lead to the conclusion.
Why learn the case? Craig says apologetics can make Christian belief seem intellectually discussable in a secular culture, steady believers facing questions, and help some skeptics take faith seriously. He urges readers to identify which premise a critic rejects and ask why, rather than trade slogans. His ideal is confidence without contempt.
What's valuable here is the distinction between argument and aggression. Treating disagreement as a shared examination of premises can lower the social cost of changing one's mind, while a combative style often turns a truth-seeking conversation into a contest of identity. The book also frames apologetics as preparation, not a guaranteed conversion technique. Its own debate stories celebrate persuasion, but a fair conversation must allow that a thoughtful person may understand an argument and still reject a premise. Respect is not merely good manners. It is part of taking another person's reasoning seriously.
Ask what disappears if God does not exist before debating evidence.
Three existential stakes. Craig argues that if God and life beyond death do not exist, life has no ultimate meaning, value, or purpose. Meaning concerns significance, value concerns good and evil, and purpose concerns goals. People can still love, choose goals, and care about justice, he says, but those commitments would be subjective rather than grounded in a reality beyond human opinion.
The question became personal. As a teenager, Craig felt alienated and empty despite attending church. A classmate's account of faith prompted him to read the New Testament and investigate Christianity. He describes finding in belief not simply comfort, but a framework in which life mattered. Nietzsche's story of the marketplace madman supplies the wider warning: people may not yet grasp the consequences of abandoning God.
The chapter's existential challenge works best as an invitation to examine assumptions, not as a proof that atheism is false. A worldview can have emotionally severe implications without being logically disproved by them. The distinction between subjective purpose and objective purpose is crucial: a secular person may regard love, work, or justice as genuinely meaningful without claiming that the universe assigned those purposes. Craig argues that this is not ultimate meaning. Whether that standard is necessary is precisely the dispute, and clarifying it makes the conversation more honest.
Trace contingent reality to an explanation that is not contingent.
Leibniz's question. The argument begins by asking why anything exists rather than nothing. Craig presents three premises: everything that exists has an explanation, the universe's explanation would be God, and the universe exists. The conclusion is that God explains the universe. An explanation may be external, as with a caused object, or internal to a thing's nature, as with something that exists necessarily.
Where the dispute sits. The universe appears to be contingent: it could have been different or absent. Craig rejects exempting it from the demand for explanation as a taxicab fallacy, an arbitrary exception made only after reaching the desired conclusion. He argues that a cause of all physical reality must transcend it, and that an abstract object cannot cause anything, leaving an unembodied mind as the explanation.
This argument usefully separates two questions that are often blurred: whether the universe has an explanation, and what sort of explanation it could have. Its force depends especially on the principle that contingent reality must be explainable. The move from a transcendent cause to a mind is another step, not simply a restatement of the first premise. Critics can also ask whether the universe, its laws, or some deeper physical reality might be necessary. The argument sharpens those options, but its conclusion is only as strong as its contested premises.
Treat the universe's beginning as evidence, not as a complete proof of God.
The kalam argument. Craig gives a three-step case: whatever begins to exist has a cause; the universe began to exist; therefore, the universe has a cause. He supports the first premise with ordinary experience and argues that something cannot arise from absolute nothing, which is not empty space but the absence of anything at all.
Two kinds of support. Philosophically, he argues that an actually infinite past would produce absurdities, illustrated by Hilbert's Hotel, an imagined infinite hotel that can be full and still make room for more guests. Scientifically, he cites cosmic expansion and thermodynamics as evidence for a finite past. He also discusses the Borde-Guth-Vilenkin result, which he says implies a past boundary for expanding universes. He infers a powerful, immaterial, personal cause.
The argument is memorable because it links an abstract question about infinity to observations about the universe. Yet a beginning does not, by itself, establish creation from nothing, identify a cause, or prove that the cause is personal. Those conclusions require additional philosophical premises. Hilbert's Hotel illustrates unusual properties of infinite sets, but whether those properties make an infinite past impossible remains debated. Likewise, cosmological theorems apply under specific physical assumptions. The strongest reading is cumulative: the scientific case motivates a metaphysical question, rather than science independently confirming every feature of the proposed Creator.
When life depends on narrow cosmic settings, compare design, necessity, and chance.
What fine-tuning means. The term describes how a narrow range of values for certain constants and initial conditions appears necessary for a life-permitting universe. It does not, by definition, mean that anyone designed the universe. Craig cites examples involving the weak nuclear force and the cosmological constant, whose values would need to fall within extraordinarily narrow ranges for life as we know it to be possible.
Three explanations. He considers physical necessity, chance, and design. He argues that known laws do not require the constants to have life-permitting values, and that chance seems inadequate given the tiny ranges involved. A multiverse, in which many universes have different settings, is offered as a possible chance-based account. Craig replies that its generating mechanism may itself need explanation and that observers can only find themselves in a life-permitting universe.
The chapter's key contribution is to turn fine-tuning into a comparison among explanations rather than a bare appeal to wonder. The distinction between a life-permitting universe and this particular universe also clarifies why the lottery analogy can mislead. Still, probability claims require a defensible range of possible settings and a measure assigning probabilities to them. A multiverse may be speculative, but the design inference also relies on assumptions about what a designer would choose. The argument is strongest as a challenge to simple appeals to luck, not as a calculation that settles the matter.
Separate whether people behave morally from what makes morality objectively binding.
Conduct versus foundation. Craig is not claiming that atheists cannot behave well. He asks whether moral values and duties are objectively true if God does not exist. Values concern what is good or bad; duties concern what is right or wrong. His moral argument is: without God, objective morality does not exist; objective moral values and duties do exist; therefore, God exists.
The grounding problem. He argues that evolutionary and social accounts can explain how moral beliefs develop, but do not establish that they are objectively binding. Against the Euthyphro dilemma, he proposes that goodness rests on God's nature, while moral obligations arise from God's commands. He also challenges secular moral realism and humanism to explain why human flourishing, rather than some other goal, has ultimate moral authority.
The book wisely distinguishes moral motivation from moral grounding. Knowing that a person can act generously without religious belief does not answer whether moral claims are true independently of belief. The harder question is whether objective moral facts require a divine source. Secular philosophers offer alternatives, including moral realism that treats some moral truths as basic rather than the products of preference. Craig argues that such accounts leave value unexplained; his own account must explain how divine commands connect to moral obligation without making morality arbitrary. The debate is substantive, not a test of anyone's character.
Ask whether suffering disproves God, or makes belief emotionally unbearable.
Two different problems. The logical problem claims that God and suffering cannot coexist. Craig argues that the claim depends on assumptions: an all-powerful God could create any world, and a loving God would prefer a world without suffering. If free choices matter, or suffering could be permitted for an overriding good, those assumptions are not necessarily true. That answers the logical challenge, though not every question about particular tragedies.
The evidential and emotional weight. The stronger challenge is that apparently pointless suffering makes God seem unlikely. Craig says human limits prevent confidently judging God's reasons, and that suffering must be weighed against the full case for God. He begins with the agonizing deaths of two children in disasters, then emphasizes that, in grief, a loving presence and the Christian image of Christ sharing suffering may matter more than argument.
The distinction between logical consistency and evidential plausibility is essential. Showing that God and suffering could coexist does not show that they do, or explain why a specific child suffered. Appeals to human limits can restrain overconfidence, but they can also become a way to protect any belief from evidence. The chapter is most humane when it recognizes that an argument cannot substitute for listening, presence, and care. Its free-will response addresses some human-caused suffering more directly than earthquakes or disease. The emotional problem deserves compassion before it receives a syllogism.
Read the Gospels as ancient sources, then test claims case by case.
Sources before conclusions. Craig treats the New Testament writings as separate ancient documents to be assessed historically, not as a single sacred volume whose authority must be assumed. He contrasts these first-century sources with later apocryphal gospels, which he regards as less useful for reconstructing Jesus. He also argues that short gaps between events and early sources, eyewitness contact, and the writers' demonstrated accuracy support taking the Gospels seriously.
Test particular claims. Craig lists historical fit, early independent sources, embarrassment, dissimilarity, Semitic features, and coherence as signs that can strengthen a claim's credibility. Applying them, he argues that Jesus understood himself as Messiah, God's unique Son, and the Son of Man. His actions, including teaching with unusual authority, forgiving sins, and proclaiming God's kingdom, reinforce that picture.
The method's most useful safeguard is its insistence that specific claims be examined rather than accepted or rejected wholesale. Multiple sources, awkward details, and historical context can increase confidence, but none functions as a mechanical authenticity detector. Source independence is often difficult to establish, and a story's embarrassment to one audience may not make it embarrassing to another. Craig's case for Jesus' self-understanding gains force from the convergence of titles, actions, and early Christian belief. Readers should distinguish that historical inference from the further theological conclusion that the claims were true.
Judge resurrection explanations by how much evidence they explain, not by preference.
Three facts to explain. Craig's historical case centers on Jesus' empty tomb, reports of appearances after death, and the disciples' early conviction that God had raised him. He argues that the burial and empty-tomb accounts are early and multiply attested, that the appearance reports include individuals and groups, and that the disciples' belief arose despite Jewish expectations that a crucified Messiah had failed.
Compare explanations. Craig assesses conspiracy, survival from crucifixion, body displacement, hallucination, and resurrection by explanatory scope, explanatory power, plausibility, and how many unsupported assumptions each requires. He argues that naturalistic alternatives explain only parts of the evidence or require extra conjectures. If God exists, he concludes, the resurrection hypothesis best accounts for the whole pattern and vindicates Jesus' claims.
The method of inference to the best explanation is more revealing than arguing over one isolated detail. A theory that explains the tomb but not the disciples' experiences is incomplete, just as a theory of visions must account for the movement's specific resurrection belief. The central difficulty is how background expectations shape the comparison. A historian who assigns miracles a very low prior probability may rank the same evidence differently from someone who already believes in God. That does not make the evidence irrelevant, but it makes the philosophical assumptions visible and worth discussing openly.
Hold Christ's uniqueness alongside the duty to reach people others never reached.
Particularism and pluralism. Christian particularism says salvation comes through Christ alone; religious pluralism says multiple religions provide paths to salvation. Craig rejects arguments that particularism is false merely because believers may be arrogant or inherit their faith from their culture. He says those attacks criticize believers or belief origins, not the truth of the claim itself.
The difficult question. What of people who never encounter Christianity? Craig argues that God may judge them through general revelation in nature and conscience, and that Christ's saving work could apply without conscious knowledge of Christ. He also offers a possible providential answer: God may arrange history so that anyone who would accept the gospel encounters it. These are possibilities, he stresses, not demonstrated accounts. For him, uncertainty strengthens rather than weakens the duty to share the gospel.
The chapter distinguishes the truth question from the moral discomfort surrounding religious exclusivity. That distinction prevents a popular accusation from settling a theological question by itself. But the proposed account of providence is a possibility argument: it shows how particularism might be coherent, not that history was in fact arranged this way. The claim also leaves a practical tension between confidence in God's ordering and urgency about mission. Craig resolves it through responsibility: Christians cannot know who would respond, so they must communicate rather than presume that another person has already received sufficient light.
Analysis
On Guard is a Christian apologetics manual, part philosophical argument, part historical case, and part personal testimony. Its architecture is cumulative: first make the stakes of belief vivid, then argue for a Creator through existence, cosmic origins, fine-tuning, and morality; next address suffering; finally move from theism to Jesus and the claim that salvation is uniquely through him. Craig's own conversion and academic journey give the book a narrative thread, while memorized arguments and anticipated objections make it a practical training text.
The strongest feature is its insistence that disagreement should be made precise. Instead of treating skepticism as one undifferentiated rejection, Craig asks which premise is disputed, what evidence supports it, and what competing explanation accounts for the facts. His treatment of the resurrection is most persuasive as a comparative argument: several historical claims are assembled, then rival explanations are tested for explanatory reach and cost. The book also correctly distinguishes emotional distress from logical contradiction, preventing a formal answer from pretending to dissolve grief.
Its chief limitation is that it often presents contested philosophical premises and selective scholarly claims with more certainty than the debates warrant. The principle that everything has an explanation, the impossibility of an actual infinite past, the interpretation of fine-tuning probabilities, and the grounding of morality in God all face sophisticated objections. Likewise, historical evidence can establish early belief and reported experiences more readily than it can establish a miracle without prior assumptions about what causes are possible. The book's scientific claims reflect its 2010 context and should not be treated as a current technical survey.
The best use of the book is therefore neither as a neutral map nor as a substitute for inquiry. It is a clear statement of one influential case, valuable for learning how its parts connect and where the real disagreements lie. Readers gain most by separating validity from soundness, possibility from probability, and historical evidence from theological interpretation.
Review Summary
On Guard presents logical arguments for Christianity and God's existence, covering topics like cosmology, morality, and Jesus' resurrection. Many readers found it an accessible introduction to apologetics, praising Craig's clear explanations and debate preparation. Some appreciated the philosophical depth, while others found parts challenging. Most reviewers recommended it for Christians seeking to defend their faith, though a few critics felt the arguments were flawed or lacked compassion.
People Also Read
Glossary
Apologetics
Reasoned defense of faithFrom the Greek apologia, meaning a defense, apologetics is the practice of presenting reasons and evidence for Christian belief and answering objections to it. Craig distinguishes this from apologizing for faith or arguing aggressively. Its purpose in the book is to equip believers, shape how Christianity is perceived, and make the case intelligible to skeptics.
Reductio ad absurdum
Expose an opposite's absurdityA form of reasoning that examines the consequences of a claim by supposing its opposite and showing that the opposite leads to an absurdity. Craig uses the idea to introduce the implications he sees in atheism, especially the loss of objective meaning, value, and purpose. He treats this as a reason to take God's existence seriously, not as proof by itself.
Necessary or Contingent
Two modes of existenceIn the book's philosophical vocabulary, a necessary thing exists by the necessity of its own nature and cannot fail to exist. A contingent thing could have failed to exist and depends on something beyond itself for an explanation. Craig uses this distinction in arguing that the universe needs an explanation and that God would be an uncaused, necessarily existing reality.
Kalam cosmological argument
Beginning implies a causeAn argument that whatever begins to exist has a cause, the universe began to exist, and therefore the universe has a cause. Craig draws on medieval Christian, Jewish, and Muslim thought and supports the beginning premise with philosophical arguments about an infinite past and scientific evidence about cosmic history. He concludes that the cause is a transcendent personal Creator.
Taxicab fallacy
Arbitrary exception to explanationCraig's label for accepting a general demand for explanation but exempting the universe once the argument reaches it. The name refers to the idea that an explanation cannot be dismissed like a cab after it has delivered someone to a desired conclusion. The objection is meant to challenge the claim that everything has an explanation except the universe.
Fine-tuning
Narrow life-permitting conditionsThe term refers to the claim that certain constants of nature and initial physical quantities must fall within narrow ranges for a life-permitting universe to exist. It is a description of the ranges, not a conclusion that the universe was designed. Craig compares design, physical necessity, and chance as possible explanations, and discusses the multiverse as a chance-based proposal.
Euthyphro dilemma
Challenge to divine moralityA challenge asking whether something is good because God commands it, which can make goodness seem arbitrary, or whether God commands it because it is already good, which can make goodness independent of God. Craig rejects this as a false choice. He argues that God's nature is the standard of goodness and that divine commands express that nature.
Logical and evidential problem of suffering
Two challenges from sufferingThe logical version claims that God and suffering cannot coexist without contradiction. The evidential version argues that the amount or apparent pointlessness of suffering makes God's existence unlikely. Craig answers the first by questioning assumptions about divine power, freedom, and possible goods, and the second by appealing to human limits, the wider evidence, and Christian beliefs about suffering and eternal life.
Criteria of authenticity
Signs supporting historical credibilityHistorical indicators used to assess particular Gospel reports, rather than automatically accept or reject a whole document. Craig lists historical fit, early independent sources, embarrassment, dissimilarity, Semitic features, and coherence. He presents them as signs that can increase the probability that an incident is historical, not as rules that independently prove a claim.
Resurrection hypothesis
God raised JesusCraig's proposed explanation of three central historical claims: Jesus' tomb was found empty, individuals and groups experienced appearances of Jesus after his death, and his followers came to believe that God had raised him. He compares this explanation with naturalistic alternatives using criteria such as explanatory scope, power, plausibility, and the number of unsupported assumptions required.
Religious particularism
One path of salvationThe view that salvation is available through one religious path. In Craig's discussion, Christian particularism holds that salvation comes through Christ alone, although he allows that Christ's saving work might apply to people without their conscious knowledge of him. It contrasts with religious pluralism, the view that multiple religions provide paths to salvation.
General revelation
Knowledge through nature and conscienceTheological term for knowledge of God said to be available to people generally through the natural world and their moral awareness. Craig contrasts it with special revelation, which comes to particular people through Scripture and, supremely, Jesus Christ. He argues that people may be judged by their response to general revelation even if they have not heard the Christian gospel.
Download PDF
Download EPUB
.epub digital book format is ideal for reading ebooks on phones, tablets, and e-readers.