How We Communicate at End-of-Life

Last year on my wife’s birthday, we were going home after dinner and admired the double set of rainbows that stretched over the gabled townhouses of our lovely city. It was a perfect gift.

When it comes to language at the end of life, most people think first about the rainbows. These are the profound final utterances a person says, expressions of self that glitter against the darkness. They’re beautiful because they’re evanescent; they’re easily carried away in stories, especially if they’re witty, unexpected, or reveal the character of a person. In certain religions, they symbolize hope and salvation, confirming a person’s goodness.

Yet they are as rare as rainbows, the result of a narrow set of special conditions. You can’t predict them, either. So it seems unfortunate that the perfectly composed, clearly articulated witticism is the cultural ideal of the end of a person’s linguistic powers. They’re not necessarily attainable. They’re certainly not natural.

For these and other reasons, I think it’s time to talk about the rain. Where I live, it rains frequently, and it’s often useful to know when it will rain and how much. What do we know about language at the end of life in cases of dying from illness or old age?

It’s useful to start with the idea of an interaction window, which is a way of thinking about our instinct to interact with each other and also the accessibility of the dying person. At the end of life, the aperture of that window gets smaller, sometimes slowly, sometimes all at once.

What do those changes look like?

One of the few studies about this showed that at five days from death, about 45% of dying cancer patients could manage complex communications, a number that dropped to 10% on the last day of life. Meanwhile, about 30% of people were incoherent five days out, a percentage that doubled to 60% on the last day. Only about a third of dying people can manage clear, simple communications until the very last day.

To me, that’s a very useful fact about the rain, with the caveat that we know much less about what people dying of other causes tend to do.

When people do make a final spoken utterance, they tend to say fixed phrases, things they learned and remembered as a group of words. They say, “I love you.” “Thank you.” They say something religious. They may also say names, and they may curse.

All this makes sense because language of this sort has been overlearned over a lifetime and is connected to emotions. Thus, it’s more easily remembered and easier to say.

Yet dying isn’t particularly talkative, which is a useful fact.

There’s a lot of silence, especially in the final days, when the dying are often medicated to keep them calm and pain-free. But silence can have many sources and meanings. Maybe people don’t want to talk. It’s also a sign of impending death. Studies show that when people stop responding to verbal stimuli, it’s a good predictor that they’ll die within three days.

Yet they may have been able to interact nonverbally for some time.

People find ways to communicate using facial expressions and touch, and there are touching stories about this, too. A palliative care doctor told me about a man who looked his young daughters intently in their eyes, one by one, as if to say, “I love you.”

To a linguist like me, it’s fascinating how readily people in changing communicative circumstances can adapt in order to stay in contact.

At the same time, there’s a lot of ambiguity in what dying people produce.

What does a moan mean? What does it mean if someone holds up a finger? They might be agitated, or they might be a devout Muslim saying a final prayer.

And what about that nonsensical speaking? They might be muttering, or they might have returned to speaking their mother tongue.

Delving into the range of what actually happens and what most people will experience, rather than only what we wish would happen based on what a few people experience, is important. It sets our expectations, and it guides our plans.

In America, it’s also important because we live in a hypercognitive society that prizes knowing things and talking about them. When language goes, we believe, the person’s not there.

But losing language and the ability to interact is a normal part of the process. The interaction window closes for everyone; it doesn’t reflect a person’s goodness or value.

So you can release yourself from the expectation that you have to say anything or be anything when you’re dying.

It’s not time for a performance.