The free will question is usually presented as a binary: either we choose freely or everything is determined. The actual landscape of positions is more varied, and the varieties matter.
The basic tension
If every event has prior causes, and our decisions are events, then our decisions have prior causes stretching back before we existed.
In which case, in what sense did we choose?
Adding randomness doesn't obviously help. An action that occurs randomly is not more free than one that was caused; it's arguably less, since randomness isn't control either.
This is the problem, and it's remarkably resistant.
The main positions
Hard determinism. Events are determined, therefore free will doesn't exist, therefore our practices of praise and blame need revision.
Libertarian free will. Genuine free choice exists, which requires that determinism is false at least for decisions. Various accounts have been offered of how this works; none is uncontroversial.
Compatibilism. The dominant position among contemporary philosophers. Free will and determinism are compatible, because what free will requires is not the absence of causes but the right kind of causes.
On this view, an action is free if it flows from your own desires and reasoning, without external compulsion. That your desires have causes doesn't make the action unfree — it makes it yours.
Hard incompatibilism. Free will is impossible whether or not determinism is true, since neither determination nor randomness provides what's required.
The theological versions
Religious traditions engaged with this extensively and independently.
The problem in that context is divine foreknowledge. If an omniscient being knows what you will do, can you do otherwise?
Various responses developed: that divine knowledge is outside time and therefore not foreknowledge in the relevant sense; that knowing what someone will freely do doesn't cause it; that God permits genuine freedom by limiting knowledge.
The predestination debates within Christian theology were fought at length and with considerable acrimony, and produced sophisticated positions on both sides.
Islamic theology developed parallel debates about divine decree and human responsibility.
Indian traditions engaged with related questions through karma — how a system of consequences for actions relates to the freedom to act.
The empirical contribution
Neuroscience has been invoked and its contribution is more limited than popular accounts suggest.
Experiments measuring brain activity before reported conscious decisions found neural signals preceding the reported moment of deciding, which was widely reported as demonstrating that decisions are made before we're aware of them.
Subsequent work has substantially complicated this. Interpretations of what the measured signal represents have been challenged, the tasks used were arbitrary and unlike meaningful decisions, and later studies suggested a capacity to veto an initiated action.
The reasonable summary: the experiments are interesting, they don't settle the philosophical question, and the confident conclusions drawn from them exceed what the data supports.
The practical question
What follows for how we treat people, which is where the argument actually matters.
If free will in the strong sense doesn't exist, does responsibility survive?
Compatibilists say yes: holding people responsible is justified because it influences behaviour, and because actions flowing from a person's own character are attributable to them.
Hard determinists frequently argue for revision rather than abandonment: retaining forward-looking practices that affect behaviour, while abandoning retributive attitudes that assume ultimate desert.
This has direct implications for criminal justice, and the argument connects to real policy debates about the purposes of punishment.
The experimental findings on belief
An empirical line worth mentioning. Studies have examined what happens when people are induced to doubt free will.
Some found increases in dishonest behaviour and reduced helping following such manipulations. These findings have had mixed replication, with several attempts finding smaller or absent effects.
Which is worth knowing both as a substantive finding and as an instance of the replication difficulties affecting this area of psychology generally.
Where it leaves things
The question remains open after a very long time, which suggests it's genuinely hard rather than that everyone has been careless.
What's changed is that the strongest positions have become harder to hold. Simple libertarian accounts struggle to explain the mechanism; simple hard determinism struggles to account for the phenomenology and for the practices that seem indispensable.
Compatibilism dominates among philosophers partly because it preserves what seems necessary while conceding what seems undeniable, which some regard as a solution and others as a redefinition.
For anyone approaching it, the useful move is probably to notice which question you're asking. Whether decisions have prior causes is a question about physics. Whether people can be held responsible is a question about practice. The two are related and they're not the same question, and a great deal of confusion comes from treating them as one.
The phenomenology
One element that arguments frequently underweight: whatever the metaphysics, the experience of deciding is extremely robust.
People who accept determinism intellectually generally continue to experience deliberation as open, to feel responsible for what they do, and to hold others responsible. The belief does not dissolve the experience.
Some philosophers treat this as evidence that something real is being tracked. Others treat it as a cognitive illusion that persists despite being seen through, comparable to visual illusions that remain compelling after explanation.
Either way it is a fact requiring accommodation. A position that implies we should stop experiencing choice is describing something nobody has managed, including its defenders.