In February, the internationally acclaimed novelist Harper Lee surprised the world with news that in July, HarperCollins will publish her second novel, Go Set a Watchman. The publisher unveiled the cover two weeks ago. Written before her masterwork To Kill a Mockingbird, this one had been buried in a drawer for decades.

Long ago, the author, certain that Mockingbird was her first and last work, decided that Watchman should not be published. Fifty years later, she changed her mind.

Some wonder, though, whether Lee is yet another older adult who's a victim of fraud and abuse. The state of Alabama is investigating. Lee is 88 and resides in an assisted-living facility. Friends say she suffers from memory, hearing, and vision loss. Her day-to-day affairs are supervised by the same lawyer who discovered the long-lost novel and negotiated its publication.

How would we know that Lee was capable of making the decision to publish a novel she long ago swore not to publish?

Cases such as hers are an immense public-health problem. Changes in older adults' cognition and need for help with daily tasks, together with accumulated lifetime wealth, make them easy prey for those who want to exploit or abuse them. And if they lose their wealth, they have limited, if any, ability to start over and recover their losses. As a result, family or the state need to step in and pay.

This case is somewhat atypical. Publishing Watchman will in fact add to Lee's already-substantial wealth (sales of Mockingbird earn her about $9,000 a day). But her case highlights an even more disturbing feature of elder abuse. It harms the person's dignity and sullies the denouement of lives otherwise well-lived and self-determined.

For Lee, publishing Watchman will reshape her carefully lived legacy. Is she, in some sense, mistakenly killing her own mockingbird?

The answer to this question engages decades of scholarship at the intersections of ethics, law, medicine, and psychology. We no longer use broad generalizations about a person, such as whether her decision was "reasonable," or whether she has dementia. Instead, capacity is grounded in an assessment of an adult's abilities to make a specific decision.

One of the essential abilities is to understand a decision's essential facts - for example, what's a novel, what's Watchman, who wrote it, and what does "publishing a novel" mean?

In my clinical practice caring for patients with Alzheimer's disease, impairments in understanding often explain why they're not able to make a decision. While they're still able to express a choice - "I do not want to go to the senior center!" - they're often unable to understand what a senior center is, its benefits and downsides.

A notable feature of this assessment is the circumscribed role of cognitive testing, such as asking the person to memorize three words and recall them later, or to draw a clock to show that it is 20 minutes to eight - both common ways to assess, respectively, memory and the ability to attend and concentrate. Cognitive tests such as these are important to raise the concern that a person may lack capacity and to explain why she lacks capacity, but they are not capacity.

This assessment doesn't judge whether the decision is what a "reasonable person" would arrive at. Harper Lee's decision reflects Harper Lee's values. But what happens when a person makes a decision based on new values, values quite different than her prior, long-held beliefs? Should we respect that decision?

This question takes a capacity assessment into difficult ethical terrain.

Classic cases include the older adult who always avoided financial risks, but who now wagers large sums at casinos, or sends bank account information to strangers to collect a share of an alleged lottery payout, or who revises a will to support a new and much younger and needy partner. Or, in the case of Lee, who swore that her first novel was her last novel, but who now has changed her mind.

People do change, and they truly can have new values. In some cases, however, these changes reflect impairments in brain function. The classic causes are conditions that damage the frontal lobes, such as from an uncommon dementia called frontotemporal lobar degeneration, or a type of traumatic brain injury.

Studies of older adults' decision-making in risky situations, or their capacity to distinguish between trustworthy and untrustworthy sources show that some older adults perform poorly on these tasks and as a result are liable to make poor decisions. While the neuroscience of these age-related changes still is being worked out, data suggest they may be explained by changes in the structure and function of frontal-lobe neurons.

Capacity cases that involve an older adult who understands the key facts but has a notable change in her values are a challenge. That they may be the result of declines in frontal-lobe function doesn't mean she is not competent. They do, however, require a competent and well-supported public-health system that can carefully assess the harms she faces from her decision and her willingness to accept them - and also whether she's the victim of excessive persuasion forcing her to act against her will. In other words, undue influence.

Jason Karlawish, M.D., is a professor of medicine, medical ethics, and health policy at the University of Pennsylvania and associate director of the Penn Memory Center.