List of Latin-script tetragraphs
Used between two velarized ("broad") consonants:
- ⟨adha⟩ and ⟨agha⟩ are used for [əi̯] (in Donegal, [eː]).
- ⟨abha⟩, ⟨amha⟩, ⟨obha⟩, ⟨odha⟩, ⟨ogha⟩ are used for [əu̯] (in Donegal, [oː]).
- ⟨omha⟩ is used for [oː].
Used between two palatalized ("slender") consonants:
- ⟨eidh⟩ and ⟨eigh⟩ are used for [əi̯].
Used between a broad and a slender consonant:
- ⟨aidh⟩ and ⟨aigh⟩ are used for [əi̯] (in Donegal, [eː]).
- ⟨oidh⟩ and ⟨oigh⟩ are used for [əi̯].
Used between a slender and a broad consonant:
- ⟨eabh⟩ and ⟨eamh⟩ are used for [əu̯] (in Donegal, [oː]).
- ⟨eadh⟩ is used for [əi̯] (in Donegal, [eː]) between a slender and a broad consonant, or for an unstressed [ə] at the end of a word.
English does not have many tetragraphs. However, when one of the elements in a sequence of digraphs is silent, such as may be are found in word-initial position in Greek or Russian loanwords, such cases might be confused with tetragraphs:
⟨chth⟩ is pronounced /θ/ or /kθ/ in chthonian and related words. When not initial, as in autochthonous, it is always pronounced /kθ/.
⟨phth⟩ is pronounced /θ/ or /fθ/ in such words as phthisis. When not initial, as in naphthol or diphthong, it is pronounced /fθ/ or with some people /pθ/.
⟨shch⟩ is used as the transcription of the Cyrillic letter Щ. It is usually read as a sequence of digraphs, /ʃ.t͡ʃ/ or /s.t͡ʃ/. However, when initial, as in shcherbakovite, the second element is silent: /ʃɜrbəˈkɒvaɪt/.
⟨ough⟩ has ten pronunciations, in half of which the digraph gh is silent. Examples are drought, bought, though, and through.
⟨illi⟩ is used to write the sound [j] in a few words such as médaillier [medaje].
In addition, trigraphs are sometimes followed by silent letters, and these sequences may be confused with tetragraphs:
⟨cque⟩ is found for [k] in words such as "grecque" and "Mecque", where the trigraph cqu is followed by the feminine suffix e.
⟨eaux⟩ is found for [o] when the silent plural suffix x is added to the trigraph eau.
⟨dcgʼ⟩ for [ᶢǀʢ]
⟨dçgʼ⟩ for [ᶢǂʢ]
⟨dqgʼ⟩ for [ᶢǃʢ]
⟨dxgʼ⟩ for [ᶢǁʢ]
⟨tsch⟩ represents [tʃ], which is a relatively uncommon phoneme in German but appears in some very common words like deutsch ("German"), Deutschland ("Germany") and tschüss ("bye").
⟨dsch⟩ represents [dʒ] in loanwords such as Dschungel ("jungle").
⟨zsch⟩ is used for [tʃ] in a few German names such as Zschopau.
There are several sequences of four letters in the Romanized Popular Alphabet that transcribe what may be single consonants, depending on the analysis. However, their pronunciations are predictable from their components. All begin with the ⟨n⟩ of prenasalization, and end with the ⟨h⟩ of aspiration. Between these is a digraph, one of ⟨dl⟩ /tˡ/, ⟨pl⟩ /pˡ/, ⟨ts⟩ /ʈ͡ʂ/, or ⟨tx⟩ /t͡s/, which may itself be predictable.
⟨ndlh⟩ is /ndˡʱ/.
⟨nplh⟩ is /mbˡʱ/.
⟨ntsh⟩ is /ɳɖʐʱ/.
⟨ntxh⟩ is /ndzʱ/.
Tetragraphs in Arrernte transcribe single consonants, but are largely predictable from their components.
⟨kngw⟩ is /ᵏŋʷ/
⟨rtnw⟩ is /ʈɳʷ/
⟨thnw⟩ and ⟨tnhw⟩ are /ᵗ̪n̪ʷ/
⟨tnyw⟩ is /ᶜɲʷ/
Piedmontese does not have tetragraphs. A hyphen may separate ⟨s⟩ from ⟨c⟩ or ⟨g⟩, when these would otherwise be read as single sounds.
⟨s-c⟩ and ⟨s-cc⟩ /stʃ/, to avoid confusion with the digraph ⟨sc⟩ for /ʃ/.
⟨s-g⟩ and ⟨s-gg⟩ are similarly used for the sequence /zdʒ/.
⟨eeuw⟩ and ⟨ieuw⟩ are used in Dutch for the sounds [eːu̯] and [iːu̯]. ⟨Uw⟩ alone stands for [yːu̯], so these sequences are not predictable.
⟨ngʼw⟩ is used for [ŋʷ] in Swahili-based alphabets. However, the apostrophe is a diacritic in Swahili, not a letter, so this is not a true tetragraph.
⟨s-ch⟩ is used in the Puter orthographic variety of the Romansh language (spoken in the Upper Engadin area in Switzerland) for the sequence /ʃtɕ/ (while the similar trigraph ⟨sch⟩ denotes the sounds /ʃ/ and /ʒ/). It is not part of the orthography of Rumantsch Grischun, but is used in place names like S-chanf and in the Puter orthography used locally in schools again since 2011.