The defenses of assumption of risk and product misuse are simply logical applications of contract notions.
In the simplest cases, the plaintiff may specifically sign a waiver of liability. In somewhat more difficult cases, the circumstances of the plaintiff’s activity and the accident may suggest an implicit agreement about who assumed the risk of an accident. For example, a player hurt in routine action in a football game, through no equipment malfunction, is generally considered to have assumed that risk, even if such assumption was not included in a specific agreement beforehand.
Product misuse is an obvious extension of this idea. For example, in International Harvester, the case in which the plaintiff read the warning on a box of bearings, understood it, and proceeded to act against it, resulting in his injury, one could justifiably say that the plaintiff had "assumed the risk" of such an accident occurring. He acted in a way which he knew created hazards, even if he did not know precisely what hazard awaited him. Any system which claims to promote individual responsibility and care would have found the plaintiff responsible for his own injuries.
Where an injury results from knowing misuse, it is the plaintiff, and not the manufacturer, who has caused the harm. That the manufacturer should be liable because the misuse was foreseeable makes no more sense than denying the plaintiff recovery for a truly defective product because the defect was foreseeable – an idea which would effectively deny any plaintiff recovery under the same definitions of foreseeability used by the courts against manufacturers.
A contractual analysis based around the distribution of risk is also the best method for solving certain thorny causation problems that will still exist.
In a simple case, say Rutherford, the question would be: Who assumes the risk that a serious injury will occur when a car hits a bridge abutment at speeds of 35 miles per hour? I believe that most auto buyers quite willingly and knowingly accept that risk, though the hypothetical of such an accident may not appear in any contract before hand. But my analysis is open to challenge. Perhaps the car was marketed as being extremely safe, indeed far safer than most cars (as is presently the case in the marketing campaign of at least one European import sedan). If so, questions might be raised as to whether the owner assumed the risk. But normally speaking, people who drive on icy roads at relatively high speeds recognize the chance of skidding, and suffering serious injury, as a result of their decision to drive. Indeed, few would expect not to be hurt in such an accident. It is a risk they assumed.
To take what is perhaps a harder example, in Eli Lilly the risk of vaginal cancer to children born of mothers who used DES was unknown, and unknowable to scientists, at the time of manufacture, and at the time the drug was taken. Since an unknowable risk cannot be guarded against, the question is whether the risk of presently unknowable future harms lies with the party producing the product or is assumed by the party choosing to try a new product. In this case, I would argue the latter, an argument made stronger because the drug carried a warning that it was an experimental drug.
A critic might argue that this is as unpredictable a test as one for negligence in a design defect. I cannot agree, but even if I did would find it preferable – at least the jury is asking the right question, which is: How did the parties voluntarily allocate the risk beforehand? Of course, had the injury occurred due to an improperly manufactured batch of the drug that had escaped inspection, or been the result of a side-effect that the manufacturer had warranted would not occur, the result would be different, since the consumer would not have received what was bargained for. But that was not the case in Eli Lilly, and should not have been the result.
The importance of reestablishing contractual allocations of risk is brought home in the search for an AIDS vaccine. Liability expert Peter Huber predicts that were a vaccine discovered, no pharmaceutical company would market it in the United States. The enormous potential liability from unknown side effects would make liability insurance unavailable, and under current tort doctrine, which ignores contract, the manufacturer could not shift this risk even to willing consumers. 
Contract, and the ability to voluntarily distribute risk, is fundamental both to individual autonomy and to assuring an available, properly functioning liability insurance market for beneficial products.