No, thank YOU, you @#$%^ machine!

I’ve heard a few folks complaining about automated messages — thank-you emails to be specific. In Auto-politeness, revisited, one of The Economist’s Johnson bloggers wrote,

Thanking is a real human response to a real event; I don’t know if it can be outsourced to a machine.

Here is a personal example from a while back. I received this email after submitting a review for a research journal. My response was neutral rather than negative. So I started thinking about why others respond negatively to this kind of message.

First off, I didn’t — and still don’t — perceive this message as a thank-you. Yes. I know it says “thank you” in two highly salient locations (the subject line and the first few words of the message). But I perceive this message as primarily informing me (that my review was received) rather than valuing me (by expressing gratitude for completing my review). If you’ve watched my tutorial on purpose, you know I find it useful to categorize the purpose of messages into four types as shown in the figure.

So how can I interpret an email that says “thank you” twice as something other than a thank-you? Because the purpose of a message can be communicated indirectly. Any native English speaker knows that “Do you know the time?” is ordinarily a request and not a question appropriately answered with “yes” or “no.” But indirectness does allow ambiguity in a way that “I request you tell me the time” does not. In this case, the indirectness is a conventional means of showing politeness. (There is a history of research in this area referred to as speech act theory. Revising Professional Writing discusses these conventions in the chapter on tone.) So I didn’t react negatively to the email because it never occurred to me to categorize it as a thank-you message.

The Economist blogger rightly attributed the problem of automated thank-yous to the lack of a human writer. Speech act theory includes a list of conditions that must be met for our words to count as a particular act. And one of the conditions our words must meet to count as an act of thanking is this: we must be sincerely grateful. That means when I say “thanks” with obvious sarcasm after a phone answering system hangs up on me, my words do not constitute an act of thanking. When a machine sends an automated “thank you,” it doesn’t constitute an act of thanking either because it cannot be sincerely grateful.

The fact is that, for the 10 years I served as editor of a research journal, I always acknowledged receipt of reviews by sending a short but personal email message. (I used readers’ first names and mentioned anything current and relevant we might share, etc.) I probably wouldn’t have if we had used the kind of automated system now commonplace among research publications. That would have been a mistake. Face-to-face communication with those reviewers was rarely possible. We were distributed across the globe. But I’m sure my personal thank-yous via email helped me develop relationships with the individuals who served as reviewers. Those relationships were important to me and to the success of the journal. A machine couldn’t have done that.

After giving this whole situation some thought, I would advise any amateur creating automated responses for unknown readers to get rid of the valuing or expressive statements: no more “welcome” or “sorry” or “thank you.” Stick with the other three purposes: informing, directing, or questioning. Then make sure there’s a human around to say “thank you” whenever possible.

Related articles

• Auto-politeness, revisited (economist.com)