More than 7,170 living languages are in use today. That fact alone explains why no single feature can define “how language works.” Some languages use tone to tell words apart. Some rely more on word order. Some carry a lot of grammar inside one verb. Some use cases, classifiers, or evidential markers. Some have long written traditions, while others are written in newer community-based orthographies or move across more than one script.
A useful pillar page on language features has to do one thing well: separate the language itself from the tools used to write it, then show how sound, grammar, and writing interact without turning the topic into a pile of labels. That is what this page does. It looks at writing systems, word order, tone, morphology, case, gender, classifiers, evidentiality, and the ways these features often cluster together in real languages.
One point matters from the start. A writing system is not the language. English, Turkish, and Yoruba all use alphabetic writing, yet they differ strongly in grammar and sound structure. Japanese uses a mixed writing system, but its syntax and morphology are not “mixed” in the same sense. Mandarin Chinese is often described through its characters first, yet many of its most important language features sit in tone, word order, and classifier use.
What Counts as a Language Feature
A language feature is any stable trait that helps describe how a language is built. Some features belong to sound. Some belong to grammar. Some belong to writing. Some sit between these levels, such as orthography, where sound, history, and literacy practice meet.
| Feature Area | What It Describes | Typical Questions |
|---|---|---|
| Writing | How language is represented visually | Alphabet, abjad, abugida, syllabary, logographic, mixed? |
| Phonology | Sound patterns and contrasts | Tone, stress, vowel inventory, syllable shape? |
| Syntax | How words are ordered in clauses and phrases | SOV, SVO, VSO, adjective before noun, prepositions or postpositions? |
| Morphology | How words change form | Agglutinating, fusional, analytic, polysynthetic tendencies? |
| Nominal and Verbal Marking | Grammar carried by nouns and verbs | Case, gender, classifiers, evidentiality, agreement? |
These areas overlap. Tone can mark both lexical contrast and grammar. Word order can signal grammatical roles in one language and discourse focus in another. Writing can reflect present-day pronunciation, older pronunciation, or a compromise between dialects. That is why good typology never treats one feature as the whole story.
Writing Systems and What They Really Encode
Writing systems are usually grouped by the unit they represent most directly. That unit may be a phoneme, a consonant, a consonant-plus-vowel pattern, a syllable, or a morpheme. In practice, real systems are often less neat than textbook labels suggest. English is alphabetic, but its spelling is far from a one-symbol-per-sound system. Arabic script is often called consonantal, yet vowel signs exist and some adapted forms represent vowels more fully. Japanese combines logographic and syllabic principles in the same sentence.
Main Types of Writing Systems
Typology usually works with five broad types, plus mixed systems.
- Alphabetic systems mainly represent phonemes. Latin, Greek, and Cyrillic belong here.
- Consonantal systems, often called abjads, give primary weight to consonants. Arabic and Hebrew are the standard examples.
- Alphasyllabic systems, often called abugidas, build each basic sign around a consonant and modify it for vowels. Many South and Southeast Asian scripts work this way.
- Syllabic systems use signs for syllables. Cherokee and Vai are classic living examples.
- Logographic systems represent morphemes rather than sounds as the main unit. Chinese characters are the best-known case.
- Mixed systems combine more than one principle in daily use. Japanese is the clearest example.
WALS highlights this typology directly and notes that the modern world is dominated by alphabetic writing, while syllabaries are much rarer and mixed systems are unusual. The same chapter also points out that Japanese stands apart because lexical morphemes are usually written logographically, while many grammatical elements are written syllabically.
Script, Orthography, and Language Are Not the Same Thing
This distinction is often missed in general articles, and it causes a lot of confusion.
- Language is the spoken or signed system used by a community.
- Script is the set of symbols, such as Latin, Arabic, Devanagari, or Han.
- Orthography is the practical spelling system used for one language in one script.
Serbian is a good example of a language that can be written in more than one script. English and French both use Latin script, but their orthographies differ a lot. Mandarin Chinese can be written with characters, yet pinyin gives it an alphabetic romanization layer. A language may also shift scripts over time without losing its identity as a language.
For a fuller topic page focused only on scripts and how they differ, see writing systems.
Direction, Spacing, and Visual Design
Writing systems differ in more than symbol type. They also vary in direction, spacing, letter connection, and the role of diacritics.
- English uses left-to-right spacing between words.
- Arabic and Hebrew run right to left.
- Traditional East Asian texts may run vertically.
- Arabic script is cursive by design, while Latin letters are usually separate.
- Many alphasyllabaries place vowel signs around the consonant, not just after it.
These traits affect reading speed, line layout, keyboard design, fonts, OCR, and educational materials. They are not small technical details. They shape how written language lives in schools, media, and software.
Why Digital Encoding Matters in 2026
A writing system becomes much easier to use at scale when it is encoded in global standards. Unicode is the central layer here. Unicode 17.0, released in September 2025, added 4,803 new characters and four new scripts, bringing the total to 159,801 encoded characters. That matters because a script needs stable encoding before it can be handled well across operating systems, search tools, fonts, browsers, databases, and mobile keyboards.
Unicode also makes an important distinction between a script and a block of characters. In text processing, software often relies on script properties rather than simple block ranges. That is why punctuation, combining marks, and mixed-script text need careful handling. For communities working on orthography development, this digital layer now matters almost as much as print design.
SIL’s orthography materials make the same practical point from another angle: a writing system is not just a chart of symbols. It has to work for readers, fit the sound system well enough, support literacy, and remain usable in real educational and community settings.
Sound Features: Tone, Stress, Syllables, and Vowels
Many readers first notice scripts, but speech features often explain far more about how a language operates. Tone is one of the clearest examples. In a tonal language, pitch helps distinguish words or grammatical forms. In a non-tonal language, pitch may still matter for intonation, but not usually for basic word identity in the same way.
Tone Is a Core Feature, Not a Side Detail
WALS divides a 527-language sample into three broad groups: 307 languages with no tones, 132 with a simple tone system, and 88 with a complex tone system. That works out to about 58.3% with no tones, 25.0% with simple tone, and 16.7% with complex tone. So tonal systems are not rare at all. They are a major part of global language diversity.
Tone systems vary in shape. Some use level contrasts such as high, mid, and low. Others use contour contrasts such as rising and falling. A language can also use tone in grammar, not just in vocabulary. WALS even notes that in at least one documented case, tone alone can signal negation.
This matters for writing, too. Tone marking is not an all-or-nothing choice. Some orthographies mark tone fully. Some mark it only where needed. Some leave it unmarked and rely on context. That choice depends on the structure of the language, literacy goals, and reader habits.
Syllable Shape Changes the Feel of a Language
Not all languages allow the same kinds of syllables. WALS groups syllable structure into simple, moderately complex, and complex. In a 486-language sample, 61 are simple, 274 moderately complex, and 151 complex. In percentage terms, that is about 12.6%, 56.4%, and 31.1%.
A language with simple syllables often favors open syllables such as CV, where a consonant is followed by a vowel. A language with complex syllables may allow dense clusters at the start or end of syllables. English belongs on the complex side, which helps explain forms like strengths. Languages with simpler syllable structure often sound more even in rhythm, though rhythm depends on more than syllable shape alone.
Vowel Systems Can Be Small or Wide
Vowel inventory size also varies less randomly than many people assume. WALS reports that the average number of vowel qualities in its sample is just under six, and that six-vowel systems account for 100 languages, or 17.8% of the sample discussed in that chapter. That figure is useful because it reminds us that “five vowels” is not a world default. It is only one common pattern among many.
Small vowel systems can still carry a lot of contrast through length, nasalization, tone, stress, or surrounding consonants. Larger vowel systems may rely on fine quality distinctions. The writing system then has to decide whether to show those distinctions directly, partly, or not at all.
Stress Is Not the Same as Tone
Stress highlights one syllable over another. Tone assigns pitch contrast to syllables or words. A language can have stress without lexical tone, lexical tone without strong stress, both together, or neither in the familiar European sense. Confusing stress with tone is one of the most common beginner errors in typology pages, and it hides real diversity.
Word Order and Why It Matters So Much
Word order is one of the first features many learners notice because it shapes the whole sentence. In typology, the basic question is the order of subject, object, and verb in ordinary transitive clauses. That produces the well-known labels SOV, SVO, VSO, VOS, OVS, and OSV.
Which Word Orders Are Most Common
WALS gives a clear global picture. In a 1,376-language sample, 564 languages are SOV and 488 are SVO. After that come 95 VSO, 25 VOS, 11 OVS, 4 OSV, and 189 with no dominant order. In percentage terms, SOV accounts for about 41.0% and SVO about 35.5% of the sample. Those two types clearly dominate worldwide, but they do not come close to covering everything.
SOV languages tend to place the verb late. Turkish and Japanese are familiar examples. SVO languages put the verb between subject and object, as English does. VSO is less common but well established in parts of the world. Very rare patterns such as OSV still exist, which is a good reminder that “rare” does not mean “impossible” or “unnatural.”
No Dominant Order Does Not Mean No Structure
Languages without one dominant order are not chaotic. WALS shows that this label covers several different realities. Some languages allow broad flexibility for discourse reasons. Others split patterns by clause type, auxiliary use, or information structure. German is a familiar example of a language where main clauses and subordinate clauses pull word order in different directions.
This is another place where many short articles oversimplify the topic. They present a language as “SVO” or “SOV” and stop there. Real usage often has a default pattern, plus special patterns for questions, focus, subordination, negation, or clitic placement.
Word Order Predicts Other Patterns More Often Than People Think
Word order does not live alone. WALS shows a strong link between object-verb order and adposition type. In the sample for adpositions, 577 languages use postpositions and 511 use prepositions, with smaller groups for inpositions, no dominant order, or no adpositions. The broader typological point is that OV languages often favor postpositions, while VO languages often favor prepositions.
The same kind of clustering shows up inside noun phrases. In WALS data on possessive order, 685 languages are Genitive-Noun, 468 are Noun-Genitive, and 96 show both orders without one clear default. For adjectives, the numbers lean the other way in the sample: 373 languages are Adjective-Noun, while 879 are Noun-Adjective, with 110 showing no dominant order and 5 using a special internally headed pattern. So not every phrase-level order lines up the same way worldwide.
That last point matters. Typological tendencies are real, but they are not iron rules. A language can be OV and still surprise you elsewhere. Good feature analysis tracks the pattern and the exceptions at the same time.
Morphology: How Much Grammar Sits Inside a Word
Morphology asks how words are built and how much grammatical meaning they carry. One older but still useful typological scale runs from isolating to agglutinating to fusional to introflexive. WALS explains this in direct terms: Chinese is the usual example of isolating structure, Turkish of agglutinative structure, Latin of fusional structure, and Modern Standard Arabic of introflexive structure.
No living language fits one label perfectly. English is more analytic than Russian, but it still has inflection. Turkish is a classic agglutinative language, yet not every suffix behaves the same way. Arabic shows nonconcatenative patterns very clearly, but it also uses linear affixation.
How Much Can One Verb Carry
WALS measures verbal synthesis by asking how many grammatical categories can be packed into a verbal word. In a 145-language sample, only 5 languages fall in the 0–1 category range, 24 in 2–3, 52 in 4–5, 31 in 6–7, 24 in 8–9, 7 in 10–11, and 2 in 12–13 categories per word.
That spread matters because it shows how misleading simple labels can be. Some languages keep grammar mostly outside the verb, using separate particles or strict word order. Others stack tense, aspect, mood, person, number, polarity, direction, or agreement inside one verbal form. A learner who sees a single long word may be looking at the work that English would spread across a full phrase.
One Meaning per Ending or Several at Once
WALS also tracks whether inflection tends to be monoexponential or cumulative. In the feature on selected inflectional formatives, 71 sample languages show monoexponential case marking, while smaller groups combine case with number, referentiality, or tense-aspect-mood, and 75 show no case in that feature set. In plain terms, some languages prefer neat one-piece grammar markers, while others let one marker carry more than one grammatical job.
This matters for both grammar description and literacy. When a writing system marks word boundaries, vowel changes, or clitics in one way rather than another, it is often responding to the structure of morphology, not just to sound.
Case, Gender, Classifiers, and Evidentiality
These features often decide how information is packaged in a sentence. They are not optional decoration. They can affect agreement, reference tracking, counting, and how speakers frame what they know.
Case Marking
Case marks the role of nouns or noun phrases. Some languages have little or no case marking and depend more on word order or adpositions. Others use larger case systems to show subject, object, location, instrument, possession, and more.
In WALS data on asymmetrical case marking, the sample includes 81 languages with no case marking, 79 with symmetrical systems, and several other types where case behaves unevenly across noun phrase types. This is a useful reminder that case is not only about how many labels a grammar book lists. It is also about where case applies and where it does not.
Gender Systems
Grammatical gender is not simply “male versus female nouns.” WALS defines gender through agreement. A language has a gender system only if noun classes trigger agreement outside the noun itself. In the WALS sample for number of genders, 145 languages have none, 50 have two, 26 have three, 12 have four, and 24 have five or more.
That distribution makes two things clear. First, many languages have no grammatical gender at all. Second, where gender exists, it can be much richer than the two-gender picture many learners inherit from school grammars. Some systems are partly semantic. Others are shaped more by form and history than by meaning.
Numeral Classifiers
Classifier systems ask speakers to sort nouns into counting classes. In languages with classifiers, the counting word may change depending on whether the item is flat, long, animate, round, grouped, or socially defined in some other way.
WALS reports 260 languages with no numeral classifiers in its sample, 62 where classifiers are optional, and 78 where they are obligatory. That means nearly one in five languages in that sample has obligatory numeral classifiers. They are especially common in East and Southeast Asia, though they are not limited to that zone.
Classifiers matter because they show that grammar can force speakers to notice kinds of entities while counting. They are also a good example of why translation across languages is rarely one-to-one.
Evidentiality
Evidentiality marks the source or basis of information. A language may signal whether something was seen, heard, inferred, or reported. Some languages leave this mostly to context and extra words. Others build it into grammar.
WALS data on coding of evidentiality show 181 languages with no grammatical evidentials in the sample, 131 using verbal affixes or clitics, 24 encoding evidentiality as part of the tense system, 65 using separate particles, 7 using modal morphemes, and 10 showing mixed behavior. So evidentiality is not a narrow niche. It is a real grammatical option used in many parts of the world.
This feature changes the feel of a sentence. A language with grammatical evidentials may ask speakers to state how they know something more explicitly than English usually does.
How Features Cluster in Real Languages
Languages are not lists of isolated traits. They are bundles of tendencies. Some combinations show up again and again. Others are less common but still perfectly stable.
| Language | Writing System | Typical Word Order | Sound and Grammar Notes |
|---|---|---|---|
| English | Alphabetic | SVO | No lexical tone, complex syllables, limited inflection compared with many other languages |
| Turkish | Alphabetic | SOV | Agglutinative profile, suffix-heavy morphology, postpositions fit the OV pattern well |
| Mandarin Chinese | Logographic in standard writing, with alphabetic romanization support | SVO | Lexical tone, strong classifier use, lighter inflection than many European languages |
| Japanese | Mixed logographic-syllabic | SOV | Postpositions, rich clause structure, writing system and syntax belong to different typological stories |
| Yoruba | Alphabetic | SVO | Tone is central, showing clearly that alphabetic script does not mean non-tonal language |
| Cherokee | Syllabary | Variable in discourse-rich settings | Shows that syllabic writing is still alive in the modern world |
This table is not here to flatten languages into labels. It is here to show that a single script type tells you very little about syntax, and a single word-order label tells you very little about sound structure. Good comparison starts with bundles of features, not single tags.
Why These Features Matter for Learning, Translation, and Technology
Language features affect far more than grammar books.
- They shape literacy design and spelling reform.
- They affect how dictionaries organize entries.
- They influence keyboard layouts, font support, and search indexing.
- They change what language learners find easy or hard.
- They guide NLP work such as tokenization, alignment, speech tools, OCR, and machine translation.
A language with rich morphology may need very different computational treatment from one that expresses similar meaning through separate function words. A script with combining marks or two-dimensional layout poses different engineering problems from one built around plain linear letters. A tonal language may need better pitch-aware tools in speech technology and careful choices in educational writing.
This is one reason recent Unicode updates matter so much. The work is not only historical or academic. It has direct consequences for daily digital use.
People Also Ask
What Is the Difference Between a Language and a Writing System
A language is the communication system used by a speech or signing community. A writing system is the visual method used to record that language. One language can use more than one script, and one script can be used by many languages. Orthography adds another layer by setting the spelling rules for one language in one script.
What Is the Most Common Word Order in the World
In large typological samples, SOV and SVO are the two most common dominant clause orders, with SOV slightly ahead. Still, many languages allow variation by clause type, discourse focus, or auxiliary structure, so the label is only a starting point.
Are Tonal Languages Harder Than Non-Tonal Languages
Not in any absolute sense. Tonal languages ask speakers and learners to pay close attention to pitch contrasts, while non-tonal languages may place more weight on other contrasts such as stress, consonant clusters, vowel reduction, or inflection. Difficulty depends a lot on the learner’s first language and on the feature being compared.
Can One Language Use More Than One Script
Yes. A language may be written in two scripts at the same time, may use one script for formal settings and another in daily life, or may shift scripts across history. Script choice can reflect education, religion, state policy, publishing habits, or digital convenience, though the language itself remains the same language.
Why Do Some Languages Use Classifiers or Evidentials
Because languages package meaning in different places. A classifier system can make counting depend on noun type. Evidential marking can make the source of information part of grammar. Languages without these systems still express those ideas, but they may do so with separate words or context instead of built-in grammatical markers.
Is Alphabetic Writing More Advanced Than Other Writing Systems
No. Different writing systems solve different representational problems in different ways. Alphabetic systems are now widespread, but mixed, syllabic, consonantal, and logographic systems are not lesser forms. They reflect different histories, different linguistic structures, and different traditions of literacy.
A Better Way to Read Language Features
The most useful habit is to read features in layers.
- Start with the writing system, but do not confuse it with the language.
- Check the sound system: tone, stress, vowels, and syllable shape.
- Look at clause order and phrase order together.
- Ask how much grammar is carried inside words.
- Then add case, gender, classifiers, evidentiality, and agreement.
That layered view gives a much truer picture than short lists of “interesting language facts.” It also matches how typology is actually used: not to rank languages, but to describe patterns, compare them carefully, and explain why one language may feel structurally close to another in one area and very far from it in another.
Once you look at languages this way, the title topic becomes much clearer. Writing systems, word order, tone, and the rest are not random features. They are the main routes through which human languages show both variety and order.
84 languages