What Makes a Language Hard to Learn?

What makes a language hard to learn?

A language is hard for a particular learner, for a particular task, under particular conditions. The main mechanisms are distance from languages the learner already knows, unfamiliar sound contrasts, a different writing system, new grammatical patterns, and access to useful instruction and practice. Change the learner's starting languages or the target skill, and the ranking can change.

This is why the U.S. State Department's language categories are informative but not universal. They publish normal course lengths for named Foreign Service speaking-and-reading objectives. They do not measure abstract difficulty for every learner, first language, purpose, or study setting.

Why is difficulty always relative to the learner?

Linguistic distance means the difference between languages in features such as vocabulary, sound patterns, and grammar. It is not one property stored inside the target language. It is a relationship between a starting repertoire and a destination.

A 2023 study by Nils Jaekel, Markus Ritter, and Julia Jaekel makes that point with school data. The researchers analysed English reading and listening among 4,551 fifth-grade students in Germany. Their measure of distance used cognate similarity: how much vocabulary with related form and meaning connected a student's reported home language with English. Greater distance was negatively associated with English reading and listening scores after the study controlled for several learner and background variables.

That is evidence about receptive English skills in one school context, not a universal equation for language difficulty. It does show why “English is easy” is incomplete. English may offer one learner many transferable resources and another far fewer.

The site's easiest-language guide and hardest-language guide are explicitly framed around English speakers for the same reason. Remove the starting language, and the comparison loses a necessary part of the question.

What do FSI language categories measure?

The current State Department Foreign Affairs Manual groups languages by “language difficulty and normal course of study.” It defines a normal course of study as the length generally required to attain the required proficiency, then gives different schedules for S-3/R-3 and other objectives. S-3/R-3 is a level 3 speaking and level 3 reading target on the Interagency Language Roundtable scale.

The National Academies' 2020 report on FSI assessment describes the training as intensive and full-time. Its learners are Foreign Service officers preparing for professional activities in embassies and consulates, and the typical training goal is ILR 3/3.

The categories therefore answer a bounded institutional question: how the Department groups course lengths for these learners, conditions, and objectives. They do not directly rank:

• a heritage learner with family exposure;
• a reader interested in one subject area;
• a traveller pursuing supported everyday interaction;
• a part-time independent learner;
• a person whose first language is not English;
• writing, which is not part of the cited S-3/R-3 label.

The language-timeline guide explains why a course schedule cannot be converted into a personal deadline without adding unsupported assumptions.

How does the sound system create difficulty?

Phonology is the way a language organises sounds and sound contrasts. A contrast matters when changing one sound for another can distinguish linguistic forms. Difficulty is not simply the number of sounds in a language; it depends on how the target system relates to categories the learner already uses.

Fred Eckman, Abdullah Elreyes, and Gregory Iverson examined second-language phonological learning through three relationships. A target contrast may contain two sounds absent from the learner's first language, one familiar and one unfamiliar sound, or two sounds that exist in the first language without functioning there as separate categories. Their article is deliberately limited to particular phonological constraints and notes that the relationship between perceiving and producing contrasts remains debated.

That scope matters in practice. A learner may hear a difference but not yet produce it reliably, or produce a rehearsed form without recognising it in varied speech. A pronunciation ranking that ignores the learner's existing sound categories and the target task is not explaining the mechanism.

For planning, assess perception and production separately. Use a named dictionary or university phonetics reference with audio for the variety being learned. Never invent an approximate pronunciation spelling and treat it as evidence.

How does a writing system affect the work?

A writing system changes the reading processes that can transfer from prior experience. Alphabetic systems represent language with letters, logographic systems use written units associated with words or meaningful elements, and alphasyllabic systems organise consonant-and-vowel information in characteristic written units. These labels describe broad types; they do not make every system inside a type identical.

Naima Mansuri and colleagues studied 1,073 adults reading English as a second language. They compared readers whose first languages used alphabetic, logographic, or alphasyllabic systems. The groups differed across English single-word accuracy, speed, and extended-reading measures, and the patterns were not identical across tasks. The authors conclude that first-language writing experience and multilingual experience constrain second-language reading.

That study examined English as the target language. It does not supply a universal ranking of scripts or prove that a different script makes an entire language harder. It supports a narrower conclusion: prior reading processes are part of the learner's starting point, and “reading difficulty” must be measured separately from speaking or listening.

Script also has several layers. Recognising a symbol, decoding a word, retrieving its meaning, following a sentence, and understanding an extended text are different performances. Completing the first layer does not complete the others.

What does morphology add?

Morphology is the structure of words and the way meaningful parts combine—for example, how a language marks grammatical information within a word. Again, difficulty lies in the relation between systems, not in a simple claim that “more endings” always means “harder.”

Frans van der Slik, Roeland van Hout, and Job Schepens analysed speaking-proficiency scores from 9,000 adult learners of Dutch. Their reduced measure compared eight morphological features. Learners whose first-language morphological configuration was less complex relative to Dutch showed more difficulty in that specific Dutch-speaking dataset.

The study does not justify ranking every language by a single morphology score. Its target was Dutch, its outcome was speaking proficiency, and its result concerned particular feature configurations relative to learners' first languages. Its value here is mechanistic: unfamiliar ways of packaging grammatical information can create work that a learner with transferable patterns may not face in the same form.

When evaluating grammar, ask which target patterns are new relative to the languages you know, which must be understood, and which must be produced accurately. Verify every claimed rule against a named grammar for the relevant standard or regional variety.

Can similarity make a language easier?

Sometimes, for some tasks. A cognate is a word that is similar in form and meaning across languages. In a 2025 Polish-English-Italian study, Małgorzata Foryś-Nogala and colleagues found that words similar across all three languages were known more accurately and confidently and used better in Italian sentences than words similar only across English and Italian.

The authors also stress that evidence about cross-linguistic influence in third-language vocabulary is still limited and sometimes conflicting. Similarity can provide a scaffold, but it can also invite an unchecked assumption. A familiar-looking word is a hypothesis until its meaning and use have been verified in an appropriate dictionary.

Transfer is also skill-specific. Shared vocabulary may assist some reading while unfamiliar sound contrasts still demand listening work. A familiar script may ease early decoding while grammar remains new. There is no honest way to add these into one permanent difficulty score without naming the intended performance.

Why do materials and access matter?

Language structure is only part of the project. Learners also need comprehensible material, reliable references, feedback, and opportunities to use the target skills. These do not change the language's phonology or morphology; they change the conditions under which the learner can work on them.

Compare two plans for the same language. One learner has a qualified teacher, recordings from the target variety, graded reading, and recurring conversation. Another has unlabelled clips, no feedback, and no way to check forms. The structural distance is the same, but the second plan has weaker evidence and fewer correction routes.

This is an editorial planning observation, not a numerical research estimate. Do not convert “more exposure” into a promised number of weeks. Instead, audit whether each target skill has suitable material and feedback.

How should you judge difficulty for yourself?

Build a profile rather than choosing a league table.

Then test a modest task in the actual language and variety you intend to learn. The point is not to produce a home-made scientific ranking. It is to identify where transfer helps, where new distinctions appear, and whether your available learning conditions match the goal.

The honest answer

Languages feel difficult when important target patterns do not transfer readily from what a learner already knows, when the intended skill adds a new processing demand, or when practice and feedback are hard to obtain. Research connects linguistic distance, sound-system relationships, writing-system experience, and morphological differences with particular learning outcomes. None of those findings makes difficulty an absolute property of a language.

Use FSI categories only for their stated Foreign Service context. For your own choice, name the starting languages, target variety, required skills, and available support. “Hard” then becomes a set of specific learning problems that can be planned for, rather than a verdict attached to millions of speakers and every possible learner.

Sources

• Jaekel, Ritter, and Jaekel, “Associations of Students' Linguistic Distance to the Language of Instruction and Classroom Composition with English Reading and Listening Skills” — accessed September 6, 2026. Used for the 4,551-student cognate-distance analysis and its bounded English reading and listening associations.
• U.S. Department of State, 3 FAM 3910 — accessed September 6, 2026. Used for the current definition and scope of the Department's language categories and normal courses of study.
• National Academies of Sciences, Engineering, and Medicine, A Principled Approach to Language Assessment: Considerations for the U.S. Foreign Service Institute — accessed September 6, 2026. Used for FSI's intensive, full-time, professional training context and typical ILR 3/3 goal.
• Eckman, Elreyes, and Iverson, “Some Principles of Second Language Phonology” — accessed September 6, 2026. Used for the three first-language-to-target-language phonological-contrast relationships and the study's own scope limits.
• Mansuri et al., “How Does One's First Language Writing Script Modulate Second Language Reading?” — accessed September 6, 2026. Used for the 1,073-adult comparison of first-language writing-system backgrounds and English reading measures.
• van der Slik, van Hout, and Schepens, “The Role of Morphological Complexity in Predicting the Learnability of an Additional Language” — accessed September 6, 2026. Used for the 9,000-adult Dutch speaking-proficiency analysis and its relative morphological-feature finding.
• Foryś-Nogala et al., “Cumulative L1–L2–L3 Lexical Similarity versus L2–L3 Lexical Similarity” — accessed September 6, 2026. Used for the Polish-English-Italian cognate result and the authors' warning about limited, conflicting L3 lexical evidence.