In 1984, Dallas, Texas, a call to the emergency services went catastrophically wrong. An elderly woman had stopped breathing, or was struggling to breathe, in her home. Her son, clearly distressed, called 911. The conversation between the caller and the dispatcher quickly spiraled out of control.

Both parties appeared to argue for several minutes over the condition of the woman. The caller, increasingly exasperated, refuses to give straight answers to many of the dispatcher’s questions. The dispatcher, equally frustrated, eventually hung up with a curt “'kay, b'bye”. Thirteen minutes later paramedics were sent to the home where the woman in question was pronounced dead.

“He didn't hear that the nurse’s questions were about helping the mother as best as possible,” says Tanya Stivers, a sociologist at the University of California Los Angeles, US. “The nurse was trying to establish some basic information about the mother. He found her questions antagonistic. Sometimes things that are transparent to one party, the nurse in this case, are not to another.”

At one point the dispatcher tried to speak to the elderly woman: “[You] can't, she's… seems like she's incoherent,” said the caller. When asked why she is incoherent, he replied: “How the hell do I know?” When chastised for cursing he said: “Well I don't care, ya stupid-ass questions you're asking.”

The incident received extensive coverage in the Dallas newspapers and on national TV news at the time. Officials expressed bewilderment as to how an emergency call could have been conducted this way.

The caller had given enough information – their address and that there was someone who had stopped breathing – to elicit a “Marine Corps response”, in the view of the Fire Surgeon for Salt Lake City Fire Department. This maximum response would include the dispatch of an ambulance, fire engine and paramedics.

After an investigation, the dispatcher in question was fired.

But how could an emergency call have gone so wrong? The failings on both parts subtly shows how the way we ask questions can have a big impact on the answers. Choosing certain words over others can help us persuade, win over and even cajole people we are talking to. By looking at similar miscommunications, and drawing on the experience of “expert talkers”, there are lessons for how we can ensure our own encounters do not end with dire consequences.

Keep talking

When dealing with persons in crisis, the strategy for police negotiators is to keep the subject talking. If you examine the recordings of these negotiations, however, you will find that “talk” should be avoided. Two researchers, Rein Sikveland and Elizabeth Stokoe, collaborated with British police to analyse recordings of conversations between persons in crisis and crisis negotiators. They saw that negotiators often use “talk” to begin a conversation: “Can we talk about how you are?” But this often gets pushed back by the person in crisis: “No, I don’t want to talk” or “It’s not genuine action, man, you’re just talking”.

Persons in crisis resist the request to talk because, as Stokoe points out, cultural idioms encourage us to put little value on “talk”. After all, “talk is cheap” and “talking the talk” is less meaningful than “walking the walk”. However, a single word substitution could be enough to save a life.

Perhaps because we do not have equivalent cultural idioms, “speak” seems to work. In real conversations between a negotiator and person in crisis, when the negotiator says “speak” (“I wanna come down and I wanna speak to you…”) they get their desired response. In some cases, the person in crisis interrupts the negotiator to begin talking. Despite being near-synonyms, one word is loaded with context that makes it ineffective in these scenarios, while the other is free of those associations.

Being willing

Mediators have also found power in words to turn around someone who is disengaged. These professional facilitators might assist in business negotiations, family grievances or disputes between neighbours. They are experts in making sure conversations reach as positive a conclusion as possible. In the UK, for example, all people in child custody disputes must first attempt to reach an agreement through a family mediation service.

Usually, an initial call with a mediator follows a set pattern. After introductions, the mediator explains how their service works. They then ask: “Does that sound like something you and the other party want to do?” To which the caller might reply: “Oh, I’m not sure the other party will ever agree to this, they are very difficult to deal with.” From a position of looking like the conversation might be shut down, the mediators can turn it around with: “Okay, but you would be willing to come in for a preliminary meeting.” “Oh of course,” replies the caller, “I was never not willing to try.”

The caller is probably not too enthused by the prospect of speaking to a mediator, and has shown that they do not really want to go down this route. Often when we do not want to do something we look to blame someone or something else. Here, they have pinned the blame on the other party – the other parent of the child.

It is effective for several reasons; it is not a question, it is a statement, and it allows the caller to frame themselves as a good person while not having to backtrack on blaming the other party.

Cultural idioms encourage us to put little value on “talk”

Usually, once one party has agreed to take part in a session, the other party follows, says Jan Coulton, chair of the College of Mediators in the UK. Coulton has been a mediator for almost 30 years, specialising in family mediation since 1998.

“We look at the way things are put by each party – the way they describe their problems – to move from negatives to positives,” says Coulton. “People in conflict can be very focused on the negatives. But when you look there are a lot of matters in common, often the children. They love the kids. We hang the pegs on the positives. There may be differences around how things are done. We may do that when we summarise. We do that in a very positive way back to them at the end.

Coulton says the principles for a positive and constructive discussion are framing your conversation in positivity, not to look for immediate resolutions – often mediation sessions will end with both parties agreeing to come back with alternative solutions to their problem – and to reach an agreement together.

Unintended consequences

Sometimes, the way that an answer to a question is phrased will also lead to unintended consequences.

Analysis of conversations between doctors and parents talking about the wellbeing of a child reveals they can often look like negotiations, even though they are not intended to be. Very specific behaviours on the part of the parent lead to a significant increase in prescriptions of antibiotics. What sets them apart from a negotiation is that the parents will not explicitly ask for antibiotics. In fact, when questioned before the appointment, they might not even say that they are hoping to receive antibiotics for their child. However, the doctor will often interpret the parents answers as an attempt to get the prescription out of them.

“They may present it with a name of an illness – they might say ‘I’m worried it could be a sinus infection’,” says Stivers, who has studied these conversations between doctors and parents. “As soon as parents are talking about infection doctors are more likely to think the parent expects antibiotics.”

The use of medical terms indicated to the doctors that the parents expected antibiotics. But, this wasn't true. Parents’ concerns were for getting the best possible treatment. Stivers describes this interaction as a good example of miscommunication between an expert and layperson. The doctor hears “infection” and immediately makes the connection with antibiotics. But for parents, that association does not exist.

“I always like to remember that doctors are people before they are doctors,” says Stivers. “We don’t like to deprive people of what they want, so it is not always about what is medically correct.”

Is there something else?

As with “speak” and “talk”, one word can make all the difference. In one study, physicians in the US were instructed to solicit extra concerns from patients making visits to their practice. One group of physicians were instructed to say “Is there anything else you want to address in the visit today?” and a second group were instructed to say “Is there something else you want to address in the visit today?”. A third group acted as a control and said nothing to solicit further concerns. In doing so, the researchers were able to test the effectiveness of the words “any-“ and “some-“ when used in open-ended questions.

The results were quite clear. “Anything” was as effective at soliciting extra concerns as saying nothing at all – 53% of patients mentioned their additional concern. Clearly, some patients did mention other ailments, but much less frequently than those who were asked the “some-“ question. In this group, 90% of patients with extra concerns raised them.

Conversation analysts, like Loughborough University’s Elizabeth Stokoe, suggest this is because the word “any-“ has a closing-down function. It tends to be used as a token gesture. Think about meetings that end with the chair asking “Any other business?”. How often are other issues raised at this point? Perhaps, with one word substitution, we would be more willing to raise extra concerns – “Is there something else you would like to raise?

The right question

Emergency service dispatchers are trained to deal with people who are experiencing “the worst day of their lives,” says Heidi Kevoe-Feldman, an associate professor in the department of communication studies at Northeastern University, in Boston, Massachusetts. Kevoe-Feldman collaborates with dispatchers taking real life emergency calls to find better ways of handling those calls. Callers are often emotional, so their answers might be incoherent. If they are unable to answer the questions, the dispatcher has to think on their feet.

Questions with “Yes” or “No” answers are very useful for getting information quickly, like whether the person in trouble is breathing. But more subtle questions, such as what colour the person looks, might need further prompting. In this case, a “menu” of possible answers is required, but it is important that the menu contains at least three options, otherwise it might lead someone towards an inaccurate answer.

Take this exchange from the fateful Dallas call described at the start of this article:

Dispatcher: “Okay is this a house or an apartment?”

Caller: “It… it is a home”.

This is a fairly innocuous question to establish whether or not to dispatch paramedics with light stretchers that can be taken up flights of stairs, or more robust stretchers that are only suitable for use in houses. The caller does not know that.

If they are unable to answer the questions, the dispatcher has to think on their feet

“The location question is so critical,” says Kevoe-Feldman. “Inside, outside, in a basement, in a car, what floor you are on. The help that you get is very dependent on how you answer this question.”

One of the problems is that callers often misinterpret why they are being asked questions in the first place.

“There were multiple things where it went wrong with the Dallas call but at the heart of it call takers ask a lot of questions and callers sometimes act like this is gate keeping,” says Stivers. “They think they need to answer these questions in a certain way to get what they want.”

It is worth remembering that the Dallas call is not representative of most calls to emergency services. It is also standard practice to tell the caller that an ambulance is on its way as soon as one has been dispatched, to allay fears they are not going to receive help, and then to continue asking further questions to get a greater level of detail.

Kevoe-Feldman says that if you want to be an expert talker when speaking to emergency services dispatchers; know your location. The best way to save a life is to say exactly where you are.

William Park is @williamhpark on Twitter.