Lying to children

A study published this month shows that school-aged children are more likely to lie to an adult if that adult had recently lied to them. The British Psychological Society’s Research Digest summarizes the study here.

Hays and Carver took school-aged (and preschool-aged) children and assigned them to one of two experimental conditions. In the first condition – the lie condition – the child was told that there was a large bowl of sweets in the experiment room when in fact there was no such bowl. On entering the candyless room, the adult admitted to having lied, explaining that they had made up the existence of sweets to get the child to come and play a game with them. In the second condition – the truth condition – the child was told that there was a fun game to play in the room which (depending on your idea of fun!) was the truth. Seated with their back to the experimenter, the child was then asked to guess the identity of a toy from an associated audio clue. When the ‘game’ got to the third toy, the experimenter was called out of the room and the child told not to peek at the toy behind them. When the experimenter re-entered the room, the child was asked if they promised to tell the truth. The child was then asked whether or not they had peeked.

Hays and Carver found that amongst school-aged children (but not preschoolers), those who had been lied to were both more likely to have peeked and were even more likely to lie about it. They concluded that their results ‘suggest that when an adult lies to a schoolage child, it increases the likelihood that they will, given the opportunity, cheat on a relatively innocuous task, and lie about having done so’.

Of course, there are limits to what we can infer from this, as the authors recognize. The study does not resolve whether the children lied because they were simply ‘copying’ the adult’s approach to truth telling, or whether they made more sophisticated judgments about what is ‘owed’ to a person who lets you down – a sort of honesty tit-for-tat. The study also does not show whether the effect would have transferred to a different person (would the children still have lied to a different person playing the game with them following someone else’s lie?) nor how long the effect lasts (if, on a second occasion, the same experimenter were to deliver on her promise of sweets, would the effect of the earlier lie be undone?). Further, it would be interesting to know whether the effect would be as strong if the context were more serious than playing a game.

Despite the limits of what can be extrapolated from the study, it still raises important ethical questions about how we engage with children. Whilst most of us probably have the intuition that it would be wrong and damaging to repeatedly lie to children, this concrete example of the (at least short-term) effects of a lie demonstrates that even small lies can influence children’s honesty. If dishonesty breeds dishonesty, then it gives us pause to think again about whether we should lie to children, and with what exceptions.

The authors of the study suggest a few reasons why adults lie to children: to control their behavior, to get them to cooperate, to control their emotions and because sometimes it is easier than providing an accurate but difficult (or, I would add, inappropriate) explanation. Some of the lies that such reasons motivate can be morally permissible: most people would agree that telling young children that babies get delivered by storks is not the sort of lie we should be concerned about. Telling a child that you’re not scared (of spiders, of the dark, etc.) when in fact you are is also a permissible sort of lie, given the pacifying effects on the child’s potential anxiety. However, telling a child that she will get sweets for cooperating and then failing to deliver does seem problematic (note that in the experiment the child was just told that there was a big bowl of sweets, which in fact did not exist – the request for cooperation was left implicit). The difference seems to be that in this case the lie serves to manipulate and the child is let down. When you tell a child that babies come from storks you simply make a false statement about the mechanics of the world. When you tell a child that you are not scared of spiders, you make a false statement about how you feel. When you tell the child that she will get sweets for cooperating and then do not give her sweets, you essentially break a promise. Absent very particular circumstances (e.g. if it is the only way to get the child to do something that will prevent her from being harmed), breaking a promise to benefit a child in some way is not going to be in their best interests.

Different sorts of lies will have different features. I have just briefly considered three different kinds of lie and there will be many other examples. One remaining question, however, is whether there might be a prudentially valuable by-product of exposing children to the kinds of lies which set them up to be let down, or reveal mal-intent. Although, in general, we do not want children to be dishonest or to grow up as liars, a sophisticated theory of mind requires an appreciation of how others can try to manipulate or conceal information through lies or half-truths. Whilst too much suspicion will lead to difficulties forming close relationships, a healthy awareness that one should not trust blindly makes a person better able to judge the intentions of others and serves to protect them from emotional or even physical harm. Given that people do lie from time to time, it is better to be prepared for this. Children also have to learn skills of tact and diplomacy, which often require lying. Thus, whilst the question of which lies are permissible and which lies are not is an important one, the further question of how we should go about teaching children that people sometimes lie (in a way that does not lead them to lie too often) is also worthy of attention. I am not at all saying that parents should betray their children as a didactic strategy – the reality is that these lessons tend to take care of themselves – but the point still remains that an appreciation of dishonesty is prudentially valuable, but that this has to be learnt through witnessing dishonesty. Indeed, the lesson that candy-promising strangers might not be trustworthy is one that many parents would be glad their child learns.

  • Facebook
  • Twitter
  • Reddit

3 Responses to Lying to children

  • Davide says:

    Just curious, is the fact you made this post exactly one month before Christmas a coincidence?

    Considering Santa Claus is a classical ‘should you lie to children’ debate…

    As for me: I think it’s quite obvious that teaching children that people will lie does not require lying to them.
    In fact I suspect that’s more of a post facto justification/rationalization than a serious argument.
    Just like teaching them that people are sometimes violent does not require being violent to them (although I’m glad to say I have not seen people use this line of thought to justify spanking and corporal puishment).

  • Carissa Veliz says:

    I wonder if adults are also more likely to lie to another adult if that adult had recently lied to them. If not, then there appears to be an interesting asymmetry in the wrongness of lying to children vs adults. In the first case, lying becomes doubly wrong: there is the wrongness of the first lie (assuming it is an unjustified lie), and then there is the wrongness of the second lie (the child’s lie). Can the adult be morally blamed for being the ’cause’ of both lies?

    • Davide says:

      It would surprise me if people did NOT feel more justified in lying to someone they knew had lied to them so…

      By the way, rereading the article, it says that (adults lie to children) ‘to control their behavior, to get them to cooperate, to control their emotions and because sometimes it is easier than providing an accurate but difficult (or, I would add, inappropriate) explanation’…
      I see no difference to the reasons adults have for lying to each other…


Subscribe Via Email