en.wikipedia.org/wiki/Turkish_grammar

This sentence states an absolute rule (“always”), but the literature does not support that. Modern descriptions of Turkish consistently describe it as canonically/head-final SOV with flexible word order, not as a language in which finite verbs must always end the clause.

Two kinds of evidence contradict the claim:

1. General descriptions of Turkish word order describe the language as flexible rather than strictly clause-final. For example, a recent PMC article describes Turkish as an “SOV (head-final) language with flexible word order,” and Hoffman's classic paper states that “Turkish has considerably freer word order than English.”
2. Postverbal material is explicitly attested in the literature. Frontiers summarizes corpus-based work by saying that “it is indeed possible to place new information in post-verbal position in Turkish.” If material can occur after the verb, then finite verbs are not always at the end of the clause.

So the accurate version would be something like: finite verbs are typically/canonically clause-final in neutral order, not always clause-final.

This sentence gives the wrong criterion for what an enclitic is. In linguistics, an enclitic is not defined by vowel harmony. It is defined by phonological dependence on a preceding word/host.

A standard dictionary definition says an enclitic is “a clitic that is attached to the end of another word” and forms an accentual unit with the preceding word. In other words, enclisis is about attachment/prosody, not about whether a vowel changes.

A Turkish example makes the distinction clear: the yes/no question particle mi/mı/mu/mü is treated as enclitic in Turkish descriptions, and it also obeys vowel harmony. But its being enclitic comes from how it leans on the preceding host, not from harmony itself. Many ordinary Turkish suffixes undergo vowel harmony without being called enclitics.

So the sentence confuses two different concepts:

• vowel harmony = a morphophonological alternation;
• enclitic = a phonologically dependent element attached to a host.

Those can coincide in Turkish, but one does not define the other.

The statement is incorrect because standard Turkish does allow consonant clusters in a syllable, especially in word-final/syllable-final position. What Turkish strongly disfavors is primarily native word-initial clusters, not every instance of “two consonants of a syllable.”

Academic descriptions of Turkish phonology explicitly refer to word-final consonant clusters. The University of Texas phonology materials discuss “dorsals occurring as the second C in a word-final consonant cluster” and contrast borrowed initial clusters with native final clusters. Another overview of Turkish phonotactics notes that native Turkish words do not begin with consonant clusters, but that Turkish allows certain two-consonant clusters within words.

So a blanket statement that “two consonants of a syllable need a vowel to be pronounced” is wrong. A more accurate formulation would be something like: native Turkish avoids word-initial consonant clusters, while consonant clusters are more restricted but possible in other positions, especially word-finally.