PACTE sees translation competence development as an ongoing, dynamic process, during which new knowledge and skills are acquired, but perhaps even more importantly, existing elements of competence are further developed, restructured and combined with each other and with new ones (Pacte Group, 2008a, 2014, 2017a, 2020). The process is visualised as a spiral to highlight the cyclical nature of development (see Figure 3). In other words, trainees continuously revisit subskills to modify and re-integrate them. Moreover, the authors stress that the subcompetencies do not develop at the same time and at the same pace, but they can compensate for each other. In addition, language pairs, translation direction, specialisation and the learning environment may influence the route and the rate of the acquisition process (Pacte Group, 2017b, 304–5).

PACTE uses the term “acceptability” for the quality of the target text and claims that there are three aspects of acceptability: meaning, function and language use. Acceptability is measured with the help of “rich points” (RPs), which are prototypical problems selected from the translation task. Solutions to RPs were rated as acceptable, semi-acceptable or non-acceptable. The research results suggest that the acceptability of direct translation shows a consistent rising tendency throughout the training period, and even after that, as professional translators’ acceptability scores were higher than those of graduate students. In general, RPs involving intentionality problems seem to cause the biggest challenges to students, and development is characterised as mixed on these types of problems. Intentionality problems are defined as “Difficulty in understanding information in the source text (intertextuality, speech acts, presuppositions, implicatures). Problems of comprehension” (Pacte Group, 2020, 110). Aspects of acceptability (meaning, function, language) show a rising tendency, too. The authors attribute development to training (p. 122) which sounds logical. Nevertheless, it should not be forgotten that the effect of engaging in translation was not controlled. In other words, we do not know how much the students would have developed if they had translated the same amount of text outside a formal training context.

Concerning knowledge about translation, findings indicate that more advanced students tend to have more dynamic views of translation. A dynamic concept of translation is defined “as a textual, interpretive, communicative and functionalist conception of translation” (PACTE, 2017, 4) as opposed to a static, linguistically oriented one. Interestingly, the PACTE research found that professional translators’ views were less dynamic than students’ views, the only exception being first-year students. There can be several possible explanations for this finding. Knowledge about translation was assessed with a questionnaire, and it is very likely that students were more aware of what the expected answers were, as they had just received education in the ideas tested. Moreover, as the Szeged research groups’ attempts to adapt the questionnaire to the Hungarian context failed, it became clear that many items did not work properly on the questionnaire.

The term “translation project” refers to a type of procedural knowledge which forms part of the strategic subcompetence. In more exact terms, it includes the ability to choose the most appropriate method for the actual translation task and to plan the process. The variable was measured by multiple-choice questions and answers were categorised again as dynamic or static, but this time a “no project” category was used, too. The PACTE group argues that the translation project undergoes what they call mixed development (rise and fall of mean values when comparing subgroups in the sample). Major increases were found between first- and second-year students and between fourth year students and graduates. Similarly to knowledge about translation, translation project values drop sharply between graduates and professionals (Pacte Group, 2020). Without going into a detailed evaluation of the research, it is worth noting that assessing procedural knowledge by a questionnaire is somewhat questionable. This may explain some of the unexpected findings. In addition, graduate students’ conformity to answers expected in an academic environment may again contribute to their advantage over professional translators.

The next variable investigated by PACTE was “the identification and solution of translation problems” and it was assumed to belong to strategic subcompetence and knowledge about translation. A questionnaire was used to collect data on six indicators of the identification and solution of translation problems. The indicators were as follows: (1) the overall difficulty of the translation task (as perceived by the respondent), (2) recognizing the prototypical translation problems previously identified by the researchers, (3) problem-solving procedures applied to solve prototypical translation problems, (4) respondents’ level of satisfaction with their own solutions, (5) characterizing the general difficulties in the text and (6) characterizing the prototypical problems in the text. Results show a complex picture of the development of this variable. The perception of the difficulty of the translation task fluctuates between the years and does not show a linear development. Problem-solving procedures seem to show an uneven developmental trajectory, too, but there is evidence that these procedures are restructured during the training. It is somewhat surprising that no progression was observed between the beginning and the end of the training on the following factors: perception of difficulty, identification of translation problems and satisfaction with proposed solutions. At the same time, major differences were found between graduates and professionals. Professionals found the task easier and identified a smaller number of translation problems. Problems of intentionality emerged as clearly more difficult for students than for professionals. Another interesting finding is that professionals are much more satisfied with their solutions than students, although the solutions themselves are of approximately the same quality. Even more striking is the lack of correlation between problem-solving indices and acceptability scores. The translation problems students identified most often were linguistic reformulation problems. The authors stress the subjectivity inherent in identifying translation problems and suggest that some of their unexpected results may be due to this subjectivity.

To examine the acquisition of and the interplay between strategic and instrumental subcompetencies, the research employed a “decision-making” variable. The data used for the analysis were the target texts and the screen recordings made during the translation process. Four different types of sequences of actions were identified in previous research (Pacte Group, 2017b) to describe translators’ decision-making processes while solving problems. These were named as (a) internal support (relying only on cognitive resources), (b) predominantly internal support (engaging in searches in sources but not using the hits), (c) predominantly external support (consulting several resources and/or combining the hits with own knowledge), and (d) simple external support (accepting the first hit in a dictionary). The main findings suggest that students tend to rely more on external support of any kind, whereas professionals are more likely to utilise some form of internal support. In addition, students’ acceptability scores were lower than those of professionals when relying on internal or predominantly internal support, indicating a lack of internal skills, competencies or the ability to use them at the same time effectively when making decisions. The evolution of automatization of decision-making processes shows an interesting pattern: it shows a decrease from the 1st year to the 3rd year of the training, then it starts to increase again, but even graduates do not reach the level of 1st year students. Professionals’ level of automatization is considerably higher than that of the students in any year.

Another variable that provided information on the development of strategic competence was “translation process efficacy”. It is defined as “an optimal relationship between solution acceptability and time” (Pacte Group, 2020, 183). In addition, “time taken at each stage” (orientation, development, revision) was also included in the study as an indicator of efficacy. As for total translation time, an inverse U-shaped development was found in that first-year students’ and graduates’ translation times were very similar, but students in between tended to spend more time on the task. Professionals were significantly faster than graduate students. It was, however, much more complicated to interpret process efficacy and its development. The PACTE group’s conclusion was that time and acceptability not only show different developmental paths, but there are different combinations of their relationship; as a result, translation competence development cannot be described by a simple “becoming faster and more accurate” process. Further research was called for by the group to clarify this complex relationship.

The last variable studied was the “use of instrumental resources”, which provided information on the development of instrumental subcompetence. The indicators of instrumental subcompetence were (a) the total time spent on searches, (b) time spent on searches at each stage, (c) the number of types of resources used and (d) the number and the variety of the searches performed (Pacte Group, 2020, 194). The results indicated that search time did not increase throughout the training, nor did the number of searches performed or the variety of the resources used. However, the distribution of the searches between the stages became more balanced by the end of the training. Moreover, the variety of the searches performed grew. Nevertheless, no correlation was found between indices of “use of instrumental resources” and acceptability.

The PACTE group’s research project stands out from other similar attempts to investigate translation competence. Its strengths include the large sample size and the complexity of the research design. The fact that the study was cross-sectional means that although some ideas can be formed about the development of translation competence, longitudinal studies are needed to check whether some outcome of the PACTE research is not the artefact of the design (e.g., fluctuating development on some factors). Methodologically, some assessment procedures can be questioned (e.g., the measurement of strategic competence), but most of the methods and ideas employed in the study can be used as springboards for further research.