If speaking a foreign language makes your chest tighten in a way that reading it never does, you are not fragile and you are not alone. What you are feeling has been a named, measured subject of research since the 1980s: foreign language anxiety.
In 1986, Elaine Horwitz and her colleagues published the paper that defined the field, along with a scale for measuring it that has been used in hundreds of studies since. Their claim was precise: this is not ordinary shyness and not general anxiety wearing a costume. It is a distinct, situation-specific anxiety that arises from the peculiar demands of speaking a language you have not mastered. They observed that no other field of study "implicates self-concept and self-expression" the way language does (Horwitz, Horwitz & Cope, 1986).
That is the heart of it. In your own language, you are a whole adult: precise, funny, quick. In the new language you can suddenly express less than you think, and every sentence risks presenting a smaller version of you to the room. The anxiety is not about grammar. It is about identity.
The 1986 paper broke the experience into three parts, and most people recognise themselves in at least one:
Communication apprehension. Fear of speaking with people, magnified because your tools are incomplete.
Fear of negative evaluation. The sense of being judged with every sentence, by teachers, colleagues, or native speakers, whether or not anyone is actually judging.
Test anxiety. The performance dread that arrives whenever speaking feels like an exam, even in an ordinary meeting.
Reviews of the research suggest that a substantial share of learners, by some estimates around one in three, experience this at a level that interferes with them (Horwitz, 2001). And it is not a harmless feeling. A meta-analysis pooling nearly a hundred studies found a consistent negative relationship between language anxiety and achievement: the more anxious learners are, the worse they perform, across countries, ages, and languages (Teimouri, Goetze & Plonsky, 2019).
Note the circle this creates. Anxiety worsens speaking, worse speaking confirms the fear, and the fear grows. Nobody studies their way out of that loop, because the loop was never about knowledge.
Two findings point the way out. First, since fear of negative evaluation is a core ingredient, practice that is itself evaluative feeds the very thing it claims to treat. Being scored and corrected at every turn rehearses the fear, not the speaking. Second, researchers studying the positive side of the picture found that enjoyment is not just the absence of anxiety but its own separate force, and it grows in settings that feel safe and personal (Dewaele & MacIntyre, 2014).
So the conditions matter as much as the hours: regular speaking, in a place where nothing evaluates you, where mistakes cost nothing and nobody is keeping score. That is not softness. It is what the mechanism of this particular anxiety demands.
It is also why UnMute refuses to grade you. Not as a marketing position, but because for this specific fear, the scoreboard is the problem.