Crying ‘Crying Wolf’: How Misfires and Mexican Engineering Expertise are Made Meaningful

ABSTRACT

On 28 July 2014, a smartphone app warned Mexicans of an earthquake that never came. Engineers took this misfire seriously, concerned that it might have a ‘cry wolf’ effect. They were concerned that people could lose confidence in the early warning system and then, the next time that alerts sounded before one of Mexico’s frequent and violent earthquakes, people might fail to take potentially life-saving action. In this article, I argue that these responses to the misfire reveal cries of ‘crying wolf’ as enactments and explorations of particular forms of responsibility integral to Mexican engineering subjectivities.

KEYWORDS:

• Mexico
• earthquake early warning
• misfire
• infrastructure
• engineering

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1 While the SkyAlert company had made a formal agreement with the Mexico City government to disseminate the earthquake early warning through a radio broadcast in 2011, dissemination through the app was not undertaken with CIRES support or supervision.

2 In some versions of the story, the false alarms and the subsequent inability of the titular shepherd boy to find help when he needs it costs him his flock of sheep. In others, he pays with his own life. On fables and the project of thinking through cultural practices with reference to animals more broadly, see Hartigan (2014).

3 While it might take minutes of steady shaking for a building to fall, non-structural elements slide or fall over more quickly. Early warnings can allow users to evacuate or take shelter before they are in any danger. Indeed, earthquake injuries are often associated with non-structural elements. For a survey of the applications of such systems, see Strauss & Allen (2016).

4 This is certainly not always the case. The tragic earthquake of 19 September 2017 originated from very close to Mexico City, so the system was only able to give city residents between 5 and 10 s of warning. Usually, though, as earthquakes tend to start hundreds of miles from the city, a minute is a conservative estimate of the lead-time that an earthquake early warning system can offer.

5 Archival research was undertaken in Mexico and the United Kingdom; primarily in the collections of Centro Nacional de Prevention de Disastres (CENAPRED), the Centro de Investigaciones y Estudios Superiores en Antropología Social (CIESAS), and at the Royal Geological Society in London.

6 As part of this research, I conducted 61 formal, semi-structured interviews. Of particular relevance to this paper are interviews with CIRES and representatives from SkyAlert as well as those of peer organisations and companies.

7 This is not an unusual state in which to encounter an infrastructure; see for example Hughes (1987; 1993).

8 This includes assessment of only the powerful earthquake of 8 September 2017, which might have killed as many as 100 people and the quake that followed on 19 September 2017, which the Mexican government estimates caused 369 additional deaths (as reported in Ureste 2017).

9 Counts vary significantly. While the Mexican state offers the official number of six thousand deaths (CENAPRED 2008) other official sources offer something closer to 20,000 (see CIRES 2018).

10 See Martínez (2017).

11 It was estimated that Mexico had roughly 33.3 million smartphone users at the time of the misfire (GSMA 2013).

12 From an interview with the author.

13 A smartphone app was limited by Mexico’s telephone infrastructure, which didn’t facilitate broadcast texts, but might still warn more people of an impending earthquake than a radio or television message that they had no way of receiving.

14 BBC (2014).

15 ‘YOU’LL GIVE ME THE SUGARS!!’ read one tweet, riffing on a classic joke from 1970s Mexican sitcom El Chavo del 8. Some tweets mentioned encounters with the traumatic pasts, and the experiences of people who still suffer from their encounter with the 1985 quake. Fright and encounters with trauma could have consequences more immediate than a ‘cry wolf’ effect: both the aforementioned ‘sugars’ (diabetes, which is popularly understood to be caused by stress and fright) and other health problems.

16 Sorensen (2000) remains an excellent summary of emergency communication best practices. Such logics and practices of preparation have been explored in Lakoff (2008); Anderson (2010).

Additional information

Funding

This work was supported by American Institute of Physics: [Grant in Aid]; Center for U.S.-Mexican Studies, University of California, San Diego: [grant number ITF: DI 13-124]; Division of Behavioral and Cognitive Sciences: [grant number 1357388]; Society for the History of Technology: [Melvin Kranzberg Dissertation Fellowship].