Jerry Walls sees the problem and tries to provide a solution in Hell: The Logic of Damnation (University of Notre Dame Press, 1992). He doesn't believe that God shuts people up in hell for eternity; rather he believes that people choose to remain in hell, even though God would gladly have them escape.
He agrees with the intellectualist presupposition of Talbott's argument: ...if the choice of hell is an intelligible one, there must be something about the subjective experience of choosing evil which can account for why some may prefer it to goodness (1992, 126). There has to be something about the choice which appears good to the person making it. Walls's answer is that the damned find something about hell preferable to heaven -- whether it is affirming their own justification in defiance of God's judgment, or preferring to wallow in sinful rebellion against God in protest, or whatever.
Now Talbott wonders: won't the experiences of hell convince these persons that they have made the wrong choice? Walls responds that persons in hell may rationalize their choices in some way. This hardens them, so that their sufferings become in a strange way more tolerable. Thus in "A Hell of a Choice: Reply to Talbott," Religious Studies 40(2) (2004), 203-216, Walls denies that hell is intolerable for the damned (2004, 212). Rather than be convinced otherwise, they harden themselves in the face of some evidence that their choice was a bad one, and they remained damned forever.
Walls supposes that he has shown that a person can freely self-impose damnation upon herself for an eternity. I am less than convinced.
Walls seemingly agrees with the intellectualist conviction behind Talbott's argument: if a person is freely to choose hell (if the choice is to be intelligible), there must be something about life in hell which the agent finds desirable, which she understands to be good. And yet at the same time, he seems to deny it when he supposes that the damned may deceive themselves into thinking that they are making the right choice. He writes: If we cannot deceive ourselves, there can be no sustained motive to choose evil, and hence no freedom so to choose (1992, 130). In this case, the will seems to act independently of the intellect, forcing upon it a false understanding of the world, thus enabling a sustained choice of evil.
Now which is it? Is intellect prior to will, or will prior to intellect?
I take it that persuasion is possible where two parties are disagreed so long as the disagreement is grounded in intellect. If Walls and I disagree about something, and intellect alone is the source of our disagreement, then we can convince each other, assuming that sufficient evidence is available. But if Walls's disagreement with me is not intellectually founded but rather caused by some perversity of will, contempt for me as a person, fear of the consequences of being wrong, or whatever, then I cannot persuade him merely with evidence. Other methods must be used.
Now when those persons who are in hell deceive themselves into thinking that they are making the right choice by remaining there, is this act of self-deception free and intelligible, or is it not?
If it is intelligible, then it must be grounded in some perception or understanding of their intellect, as Walls himself grants. In this case, the difference of opinion between God and the damned is an intellectual matter, and so he can convince them otherwise by providing adequate evidence that they are wrong. But if it is not intelligible, if their act of self-deception is not free, then certainly neither can their choice to remain in hell on the basis of that self-deception be free.
In this way, I think Walls's project fails. Either the self-deception of the damn is due to intellectual error of some sort, in which case they can be corrected through persuasion and proof, or else it is not, in which case it is not free self-deception and their choice to remain is also unfree.