“Advise When Ready for Game Plan”: Military Target Language Use – The Case of Directing Close Air Support

Assessing Professional Language Constructs

The literature on testing English for Specific Purposes (ESP), or Language for Specific Purposes (LSP), relevant to the JTAC work context, has flourished since the field’s inception in the 1960s (Dudley-Evans & St John, 1998). Testing ESP embodies the testing of constructs that differ from general language proficiency constructs to some extent, being influenced by the linguistic requirements of particular professional domains (Widdowson, 1998). Due to the respective differences between using general language and using professional language within a specific-content domain (Davies, 2001; Douglas, 2000; Knoch & Macqueen, 2020), there is a broad scholarly consensus that authenticity takes on particular significance when developing LSP tests (Bachmann & Palmer, 1996; Hutchinson & Waters, 1987; Davies, 2001; Douglas, 2000; Knoch & Macqueen, 2020; Lewkowicz, 2000; Moder & Halleck, 2021). Accordingly, to ensure the right construct is targeted, LSP test designers must carefully consider the language tasks, the vocabulary, and the linguistic functions needed in real-life TLU situations before embarking on the process of developing tests (Brunfaut, 2014; Moder & Halleck, 2021).

A lack of analysis of TLU might lead to the specific-purpose language domain being underexamined in the test, thereby impairing the validity of inferences made about candidates’ future work performance (Widdowson, 1998; Fulcher & Davidson, 2007). To assess what these test-takers can achieve using their second language in specific-purpose contexts (Moder & Halleck, 2021), ESP test design and content must focus on the linguistic skills candidates will need in the TLU situation (Brunfaut, 2014; Douglas, 2000; Moder & Halleck, 2021). A general language proficiency test might not be the best tool for establishing this, possibly presenting part of the reason why a renewed focus on the misuse of language tests as part of validity theory has been proposed (Carlsen & Rocca, 2021). This raises the question of whether it is valid and fair to assess military-specific language users against a general proficiency scale.

Language for Military Purposes

While peer-reviewed publications on language teaching in the military context remain few and far between (Siegel et al., 2024), English proficiency has always constituted an important component of NATO allied interoperability (Monaghan, 2012). A comprehensive study of language challenges in NATO operations, however, concluded the STANAG level descriptors were an inadequate tool for linguists (Jones & Askew, 2014); indeed, balancing the development of general language proficiency tests with some inclusion of specific-purpose language used in the military professional domain continues to present challenges for several NATO language-testing teams (Green & Wall, 2005). This is problematic since individual approaches to test design and assessment in NATO member and partner nations might not only result in introducing construct-irrelevant variance (Messick, 1996): they might jeopardize language standardization in the alliance altogether.

Construct-irrelevant variance results from including topical content and eliciting linguistic functions in the test that go beyond what stakeholders are interested in having measured (Frey, 2018). Therefore, the importance of focusing on construct, authentically testing TLU, is key (Bachmann & Palmer, 1996; Lewkowitz, 2000) when targeting communicative contexts and interaction requiring language skills other than those used in, say, general or academic contexts (Douglas, 2000; Joyce & Thomson, 2015; O’Sullivan, 2012). When testing JTACs in accordance with STANAG 6001, then, language-testing teams run the risk of extrapolating how well they (or other military specialist personnel) will perform their future work from results derived from a test designed to measure a different construct (Grande et al., 2022). If valid predictions about future work performance are to be made, tests need to accurately mirror authentic TLU (Moder & Halleck, 2021), but when testing JTACs – what, precisely, might that entail? Alternatively, which STANAG 6001 general language proficiency level does JTAC work require? Do JTACs use mostly functional language, or is advanced proficiency needed?

Although the relationship between air operations and ground operations has been established for decades, little research has been published on JTACs’ TLU (Pirnie & Vick, 2005) – despite JTACs speaking English as their second language apparently struggling throughout the alliance. Even experienced JTACs might not meet the NATO language requirement when their certification needs to be renewed (Kildevang-Jakobsen, 2019). This lends support to a study finding little alignment between STANAG-level descriptors and the spoken language tasks JTACs perform within their professional domain (Di Biase & Gratton, 2015). Thus, an operational gap exists between assessment method and TLU (Leishman, 2019).

A case study including more military content in the STANAG 6001 oral proficiency test for JTACs found that situating tasks in military-related contexts did not affect the quality of test-taker output (Grande et al., 2022). Although the study concluded JTACs use ESP in unique ways, no attempt was made to identify specific TLU characteristics, presenting a gap in our knowledge of what makes JTAC language use differ from general English language use. Where Di Biase and Gratton (2015) contended there is misalignment between general language proficiency as described in NATO STANAG 6001 and JTAC TLU, then, how is this misalignment manifested?

In 2021, the Bureau for International Language Co-ordination (BILC), custodians of STANAG 6001 and NATO’s advisory body on questions related to language training and testing, formed a working group tasked with developing a prototype JTAC language test using JTAC-specific contexts and tasks (Garza, 2025). After consulting subject-matter experts from three nations, observing JTAC simulation training and certification, and observing exercise Adriatic Strike with 26 participating nations, the working group found support for Di Biase and Gratton’s (2015) claim of a disconnect between what JTACs use English for, and the way their language skills are tested (Garza, 2025). At the time of writing, the working group has trialled integrated skills performance test tasks and been expanded to include language professionals from ten NATO member nations. It is planned for the coming JTAC language test to be used as an instrument for screening the English-language skills of candidates interested in starting JTAC training (Kildevang-Jakobsen, 2025). However, no timeline for test development completion, validation, or implementation has yet been communicated.

Studies from the civilian aviation context have argued a high level of general English language skills contributes to keeping travellers and aircraft personnel safe (Emery, 2014; Park, 2020) since standard phraseology cannot cover all possible scenarios (Drayton & Coxhead, 2023). Furthermore, the pilot and the air traffic controller – the JTAC’s civilian counterpart – depend on each other, not only in terms of producing accurate spoken language but with regard to their respective levels of listening comprehension (Garcia & Fox, 2020). Garcia and Fox’s (2020) findings thus suggest a multimodal test assessing interactional competence (Roever & Kasper, 2018) – listening, understanding, and speaking – would be as relevant for JTACs as for the civilian aviation context (Treadaway & Read, 2025). Due to the international work environment in civilian aviation, it has been argued that English as a lingua franca (ELF) is the norm rather than the exception (Prado & Tosquil Lucks, 2019), placing other demands on communicative and interactional abilities than those described in general language proficiency descriptors (Harding & McNamara, 2018).

Estival and Pennycook (2024), however, claim there are clear distinctions between aviation English and ELF. Moreover, an experimental study has suggested that working in environments characterized by high information density and a degree of noise will actively disadvantage even aviation personnel speaking their first language (Estival & Molesworth, 2020). For second-language speakers, the need to successfully negotiate meaning will almost certainly appear.

A study of the communication between an ATC and a pilot, to some extent mirroring the JTAC TLU situation, found that the spoken language construct in this context comprised solid professional knowledge in combination with the use of the correct phraseology and procedures (Kim, 2018). Another study nevertheless contended the reliance on standard phraseology could backfire (Estival et al., 2023), arguing ATCs would need general English proficiency to resolve unexpected developments (Emery, 2014), which might also apply to directing CAS. From the perspectives of teaching, testing, and assessment, it would seem establishing how JTACs use English spoken language at work is highly relevant.

This study contributes to the scholarly literature on JTAC TLU, building on the research conducted by Di Biase and Gratton (2015) and Grande and colleagues (2022), bridging the identified gap in knowledge regarding the linguistic differences between JTAC TLU and general spoken English. Using a framework synthesizing subject-matter expert opinions with lexical analyses of JTAC discourse, our study documents Norwegian JTACs’ perceptions of the balance between functional and advanced general English when directing CAS, the role played by domain-specific lexis when directing CAS, and the linguistic features separating JTAC TLU, from general English proficiency as constructs.

The study was guided by three research questions (RQs):

• RQ1: What is the balance between functional and advanced general English in JTAC TLU?

• RQ2: To what extent do JTACs rely on ESP when directing CAS?

• RQ3: Comparing transcribed examples of JTAC language output when directing CAS to transcribed samples of authentic OPI output, which linguistic features, if any, separate JTAC TLU from general language use?

Participant Sample

The questionnaire respondents (N = 30) were initially approached via the unit responsible for training JTACs in Norway, who also collected email responses, guaranteeing test-taker confidentiality. By filling in and returning the questionnaire, respondents gave their consent to taking part in the study. All participants were qualified JTACs, 67% of them still operational. 53% of the sample had 1–5 years’ experience in this position, 13% had 6–8 years’ experience, and the remaining 34% had more than nine years’ experience. 73% had been deployed. 72% were army personnel; the rest were special operations forces (SOF) personnel. 30% were officers (OF), the rest belonged to other ranks (OR). All respondents were males speaking Norwegian as their first language. 73% of the sample had been Norwegian Armed Forces employees for more than nine years.

Instruments

Questionnaire

Text Comparisons

RQ3: Comparing Transcribed Examples of JTAC Language Output When Directing CAS to Transcribed Samples of Authentic OPI Output, Which Linguistic Features, If Any, Separate JTAC TLU From General Language Use?

Basic text properties were calculated before the lexical analyses of spoken texts were conducted. For word count and number of complex (three or more syllables) words, means and medians were comparable, although CAS texts were somewhat lengthier than OPI texts. Sentence count showed a marked difference between directing CAS and OPI output (Table 6).

Table 6

Comparison of basic text properties.

TEXT PROPERTY	TOTAL	MEAN	S.D.	Mdn	N =
Word count
JTAC output	1,588	226.9	87.8	209.0	7
OPI output	1,383	197.6	45.9	218.0	7
Sentence count
JTAC output	175	25.0	9.9	25.0	7
OPI output	41	5.9	2.0	6.0	7
Complex word count
JTAC output	157	22.4	10.9	20.0	7
OPI output	156	22.3	6.7	21.0	7

Average sentence length was calculated by dividing total words by total sentences, prior to calculating the average number of complex words per sentence. CAS texts yielded an average sentence length of 9.1 words, compared to 33.7 words for OPI texts. Average number of complex words per sentence was lower for CAS texts (M = .79, SD = .34) than for OPI texts (M = 4.26, SD = 1.85). A non-parametric independent samples Mann-Whitney U test compared the use of complex words in CAS and OPI output. Results showed the null hypothesis had to be rejected: there was a statistically significant difference between the number of complex words per sentence in CAS and OPI output respectively (U = 49.00, z = 3.13, p = .001). Effect size value (d = .836) suggested high practical significance. An independent samples median test similarly showed a statistically significant difference between CAS and OPI sentence length (Figure 1).

Figure 1

Lexical analyses standardized per text length showed that when directing CAS, JTACs produce lexically dense sentences (sentences rich in content words), containing many nouns but fewer adverbs and prepositions than that found in OPI output. For pronouns and auxiliary verbs (also known as “helping verbs”), the difference in number was quite striking (Table 7).

Table 7

Lexical Analysis.

TEXT PROPERTY	GENRE	MEAN	S.D.	Mdn	N =
Lexical density	JTAC	61.3	2.5	63.2	7
full text/output	OPI	45.7	2.7	46.2	7
Nouns	JTAC	37.7	1.6	37.3	7
	OPI	19.2	5.3	18.5	7
Adjectives	JTAC	9.5	2.3	8.8	7
	OPI	7.3	2.1	7.8	7
Verbs	JTAC	10.9	0.8	10.8	7
	OPI	11.5	3.0	13.2	7
Adverbs	JTAC	3.8	1.7	3.4	7
	OPI	7.7	4.8	7.4	7
Prepositions	JTAC	9.9	2.6	9.8	7
	OPI	13.5	1.9	13.3	7
Pronouns	JTAC	1.2	0.6	1.0	7
	OPI	7.9	2.4	8.8	7
Auxiliary verbs	JTAC	2.2	0.6	2.2	7
	OPI	11.4	1.0	11.0	7

The above numbers show lexical density and word-class distributions, without documenting use of abbreviations, brevity words, standard phraseology, and military terminology. A transcribed example from Joint Publication 3-09.3 (U.S. Joint Chiefs of Staff, 2014) of JTAC turns when directing CAS attests the role of ESP in JTAC discourse. Pilots’ turns have been edited out (Table 8).

Table 8

Appendix E of JP3-9.3 (2014): Examples of Close Air Support Missions.

	EXAMPLE 3 – TYPE 2 CONTROL, BOC MISSION	LEXICAL DENSITY
1	“Keyhole in effect, Echo point NB 234 876.	62.5%
2	Proceed to Bravo 8, block 17–19, you are the only aircraft on station, advise when ready for game plan.”	52.6%
3	“Rumble 41, this is Redman 01, Type 2 control, bomb on coordinate, 1 by GBU-32 each, airburst fuzing, simultaneous impacts, advise when ready to copy 9-line.”	53.9%
4	“Bravo 8, Elevation, 1650 feet, company of infantry in trench line, NB 234 876.”	57.1%
5	“No mark, South 1100, egress right pull, back to Bravo 8, block 17–19.	61.5%
6	Advise when ready for remarks.”	60%
7	“Final attack headings 280 through 320.	50%
8	Request IP inbound and IN with heading calls.”	62.5%
9	“Rumble 41, good readback.	75%
10	Rumble 42, go with readbacks.”	60%
11	“Rumble 42, good readback; Rumble flight, report IP inbound, TOT 35, expect clearance as a flight.”	75%
12	“CONTINUE.”	100%
13	“Rumble 41 flight, CLEARED HOT.”	80%
14	“Good impacts, standby BDA.”	100%
15	“Ground commander’s intent met, enemy neutralized.	100%
16	Switch Blacklist for routing and further tasking, you are still the only aircraft on station.”	60%

In the above example, the JTAC, “Redman 01”, conducts a Type 2 control, bomb on coordinate. Unfortunately, no information about time – pacing, or words per minute – is provided: this would have been interesting from the perspectives of fluency and comprehensibility. When directing the aircraft, “Rumble 41” and “Rumble 42”, the JTAC produces 11 turns,16 sentences in total. Three sentences (12, 14, and 15) are 100% lexically dense, containing only content words. Lexical density full text is 63.2%, which is considered high, and the total number of words is 155.

Abbreviations are used five times (sentences 3, 8, 11, and 14). GBU means “guided bomb unit”; IP means “initial point”; TOT means “time on target”, and BDA means “battle damage assessment”. Brevity words (sentences 8, 12, and 13) include “IN” (direction; sentence 8); “CONTINUE” (continue present manoeuvre, does not imply clearance to engage or expend ordnance; sentence 12); and “CLEARED HOT” (“Type 1 and 2 close air support terminal control clearance to release ordnance on this pass”; sentence 13).²

The 9-liner (sentences 4 and 5) follows set protocol, presenting information in a particular order so the listener will know what to listen for, and using collocations: “Advise when ready for game plan” (sentence 2); “Advise when ready to copy 9-line” (sentence 3); and “Advise when ready for remarks” (sentence 6). These collocations serve as discourse markers: the JTAC will now be awaiting the pilot’s response. General military expressions, in addition to obvious ones such as “trench line” (sentence 4) and “commander’s intent” (sentence 15), are sometimes hidden in plain language. Chances are, for example, that “Keyhole in effect” and “Echo point” (sentence 1) might not be understood without knowing the professional context.

The lexical analyses show JTACs produce short and lexically dense sentences, using more nouns and to some extent adjectives, but fewer verbs, adverbs, prepositions, pronouns, and auxiliary verbs. Specific-purpose language (abbreviations and brevity words) constitutes a sizeable portion of the language output, facilitating pilot comprehension.

RQ1: The Balance Between Functional Language and Advanced General Language Use in JTAC Work

Our JTAC sample doubted whether a functional level (STANAG 6001 L2) of general proficiency would suffice in the JTAC profession. The SMEs agreed that military domain-specific terminology is used more in the TLU situation than general language, which could be taken to suggest a functional level of English to be sufficient, lending support to previous research finding little alignment between STANAG descriptors for L3 and the language tasks JTACs perform (Di Biase & Gratton, 2015; Garza, 2025; Grande et al., 2022).

Indeed, certain language functions rewarded in a general proficiency test may matter little when directing CAS: because both JTAC and pilot are highly proficient military professionals (Kim, 2018), communication remains effective even without, say, grammatical accuracy. Our lexical analysis supports this. Accordingly, whether a JTAC’s general language proficiency is functional or advanced may make little practical difference when directing CAS.

Nonetheless, the NATO requirement of L3 general English proficiency was not considered unreasonable by all respondents, lending support to Emery (2014), Drayton and Coxhead (2023), and Park (2020) who all contended standard phraseology might not cover all eventualities, meaning general language would be necessary. The sample confirmed the situation can change fast in the operational theatre, increasing cognitive load (Estival & Molesworth, 2020; Treadaway & Read, 2025) and requiring language adaptation. The question remains whether JTACs “adapting language” might simply involve switching protocol and continue using standard phraseology. One comment suggested more advanced English spoken-language skills are needed in the joint planning and target development stages of the operation than when directing CAS. Nonetheless, it is not unwarranted to argue that even when functioning as advisers to the ground force commander, JTACs might take a solution-oriented, narrative-descriptive, functional-language approach rather than switching to STANAG L3 advanced, abstract English.

JTACs’ need for spoken accuracy and precision (Di Biase & Gratton, 2015) was confirmed by our respondents. The pilot needs to spot the right building, know its exact location, receive and act on the correct target coordinates. This leaves no room for inaccuracies or mistakes on the JTAC’s part, and the importance of good listening comprehension skills should thus not be underestimated (Garcia & Fox, 2020); indeed, they should perhaps even be integrated in aviation spoken language assessment (Treadaway & Read, 2025). Good pacing and clear pronunciation, skills which can be trained without possessing NATO STANAG L3 general language proficiency, contribute positively to listener understanding. Moreover, the STANAG L3 descriptors state “pronunciation may be obviously foreign” without describing minimum adequate performance. This raises the question of exactly how foreign a JTAC’s pronunciation can be, without compromising accuracy, precision, and safety. Based on the questionnaire responses, the balance between functional and advanced general English use could not be firmly established (RQ1).

RQ2: The Role of English for Specific Purposes (ESP) to JTAC TLU

When consulted on the use of ESP (RQ3), respondents supported Kim’s (2018) assertion that professional context and domain-specific terminology play complementary roles. When directing CAS, protocol and standard phraseology help prevent misunderstanding, even if there is interference or static, or the pilot speaks with a heavy accent (Kim, 2018). The question remains whether adding words or deviating from standard phraseology might cause delays or a breakdown in communication (Estival et al., 2023), clearly undesirable in the JTAC work context. The CAS transcript showed how the JTAC used context-specific abbreviations, brevity words, and phraseology to produce a focused message, again suggesting a general proficiency test might be suboptimal for assessing JTACs’ spoken language skills.

This resonates with how the JTAC community has questioned the relevance of the present NATO spoken-language requirement (Grande et al., 2022) and might contribute to explaining why JTACs speaking English as their second language struggle to be rated at the level needed for certification (Kildevang-Jacobsen, 2019). Since tests have not been developed based on analyses of authentic TLU (Brunfaut, 2014; Moder & Halleck, 2021), construct-irrelevant variance could be involved (Messick, 1996), making the inferences made from test results not serving their intended purpose. Although not all respondents considered directing CAS their most critical TLU task, findings nevertheless suggest that JTAC spoken TLU is largely military specific and relies on ESP (RQ2).

RQ3: Comparing Transcribed Examples of JTAC Language Output When Directing CAS to Transcribed Samples of Authentic OPI Output, Which Linguistic Features, If Any, Separate JTAC TLU From General Language Use?

We now consider language features separating JTAC TLU from general language use (RQ3).

A reliance on a predefined format and the use of ESP in the form of abbreviations, brevity expressions, and standard phraseology has already been mentioned. Lexical analyses comparing transcribed examples of CAS output to authentic general language OPI output showed JTACs produce short, lexically dense sentences, containing more content words, and fewer adverbs, prepositions, pronouns, and auxiliaries than OPI output. Furthermore, a statistically significant difference was found between the number of complex words per sentence in CAS and OPI output, respectively, possibly explained by the need for precision, using words sparingly and accurately, when directing CAS (Di Biase & Gratton, 2015).

Moreover, it seems CAS output would be hard to follow for listeners without knowledge of JTACs’ professional work (Kim, 2018). Our findings lend support to claims made that there is a misalignment between current testing practices in accordance with STANAG 6001 and what JTACs use English language for in their profession (Di Biase & Gratton, 2015; Garza, 2025; Grande et al., 2022, Kildevang-Jakobsen, 2019). The combination of predefined format, use of domain-specific terminology, and short, lexically dense sentences makes JTAC TLU distinct from normal spoken English language (RQ3).