Unintended consequences of translation technologies: from project managers’ perspectives

ABSTRACT

Recent years have seen the advance of increasingly efficient translation and translation-related technologies, such as neural machine translation and crowdsourcing-style translator procurement platforms. These artificial-intelligence, big-data and algorithm-driven online systems are hailed as successes in the media- and technology-vendor-led public discourse. However, in light of the notion of ‘solutionism’, there may be a risk that unintended adverse consequences of these technologies on users remain obscured. As a result, a correct assessment of the influence of technologies on human actors may become difficult. In order to identify such unintended consequences of translation technologies, the present article explores technology users’ perceptions about how the technology is affecting their business practice. The discussion draws on data collected in a focus group study with 16 translation project managers. The study reveals that project managers are feeling a high level of uncertainty and unease about the effects of technology when they talk about business practices, particularly in the following areas: translators’ use of machine translation, pricing for post-editing, post-editors’ profiles and skills and technology-induced new power struggles in the industry.

KEYWORDS:

• Translation technology
• focus group
• language service providers
• project managers
• machine translation
• paid crowdsourcing

Acknowledgments

I would like to extend my gratitude to the study participants for taking time out of their busy schedules to share their insights. This study was conducted as part of the larger project, When Translation Meets Technologies: Language Service Providers (LSPs) in the Digital Age, funded by the University of Portsmouth.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes on contributor

Akiko Sakamoto is senior lecturer in translation studies and Japanese language at the University of Portsmouth, UK. She holds a PhD in translation studies from the University of Leicester, UK. Her research interests include sociology of translation, particularly the influence of technology on translation practitioners and users, pedagogy of translation and translation theories. She previously worked as an English–Japanese translator.

Notes

1 STS is an interdisciplinary field of study conducted by sociologists, philosophers, historians and anthropologists. It examines the interactions between and mutual influence of technology and human agency and behaviours. For the history and terrain of STS, see, for example, Sismondo (2010).

2 For recent work examining PMs, see Olohan and Davitti (2015).

3 The outcomes of this section are based on the following categories generated in the analysis: ‘PMs’ positive attitudes with proviso’ (35 counts across the 4 groups); ‘Whether translators use MT or not’ (11 counts/3 groups); ‘Company policy about translators’ MT use’ (8 counts/3 groups); and ‘Uncertainty about MT use’ (8 counts/3 groups), all for Q1.

4 The outcomes of this section are based on the category ‘Cost and pricing’ for Q1 (21 counts/4 groups).

5 The outcomes of this section are based on the category ‘Normal translator vs post-editor’ for Q1 (18 counts/4 groups).

6 It is also important to add that two participants provided positive comments about post-editing. One said seeing a MT output before translating is like having ‘that extra bit of energy’ and he ‘find[s that] fun’ (G2-1). The idea that post-editing is boring may be a simplistic generalisation that deserves further testing.

7 The outcomes of this section are based on the categories ‘Technology companies’ (for Q1), ‘Technology companies vs small companies’ (Q5) and ‘Big and small companies’ (Q6) (total 10 counts/4 groups) as well as the categories ‘Negative’, ‘Positive’, ‘Control’, ‘Price’, ‘World trend’ and ‘Image of translation’ for Q6 (98 counts/4 groups).

8 For the purpose of the discussions, the moderators defined the term as a ‘human translation service with near-simultaneous turnaround’, using a quote of the founder of Stepes (Stepes, 2016), and gave an example of Gengo, which claims to have on their platform ‘over 15,000 native speakers – located in 140+ countries – translated more than 130 million words across 35+ languages’ (“Gengo,”, n.d.).

9 The views reported here are those of the participants and not intended to inform readers of any merits or disadvantages of the business model itself.