How to Practise Speaking a New Language on Your Own

How to practise speaking a language alone is a practical question for anyone without a regular conversation partner, and the honest answer has two halves. Some parts of speaking can be practised on your own, and there is research suggesting it is worth doing. Other parts need another person, and no solo routine replaces them. This guide sets out what the research says, with its limits, and then a concrete 20-minute session built from it. The examples use Italian, but nothing in the routine is specific to it.

Why speaking to yourself counts as practice

The case for solo speaking rests mainly on one idea in second-language research: the comprehensible output hypothesis, developed by Merrill Swain. Wikipedia's summary says it holds that learning can take place when learners meet a gap in their knowledge of the language, notice it, and modify what they say. Swain describes three functions of output:

• Noticing: trying to speak shows you the gap between what you want to say and what you can say.
• Hypothesis testing: each sentence you produce tests a guess about the grammar, and feedback from a listener confirms or corrects it.
• Metalinguistic: reflecting on the language you produce helps you control and internalise it.

The first and third of these can happen alone. The second, as Swain frames it, needs an interlocutor.

The hypothesis is not settled. Wikipedia records Stephen Krashen's criticism that output is rare and comprehensible output rarer, and his argument that more comprehensible input is a better strategy than more output. It also notes that students in one survey placed speaking at the top of the list of what made them anxious in foreign-language learning, and that pushing learners to speak may make them uncomfortable and so hamper acquisition. Solo practice is one way to speak without an audience, but it is fair to say the field is divided on how much output matters.

What research on private speech suggests

Speaking to yourself has a name in psychology: private speech, which Wikipedia defines as speech spoken to oneself, audible but neither intended for nor directed at others. It was first studied by Vygotsky and Piaget in children. Wikipedia notes that in a self-reported questionnaire young adults also reported high levels of private speech, particularly during tasks with cognitive, memory and attention components.

For second-language learners there is some direct evidence. A 2025 study by Jiang and colleagues in Brain Sciences asked 64 Chinese-speaking learners of English to complete a picture-description task, and reported that private speech facilitated their oral production. In a second experiment, brain imaging of 32 learners who naturally used private speech found it produced connectivity patterns similar to inner speech. The authors suggest its role changes with proficiency, supporting basic language processing in lower-proficiency learners.

That is one study, with one task and one learner group, so it supports trying private speech rather than proving it works for everyone.

Reading aloud and shadowing

Two more structured forms of solo speaking have been studied. A 2021 study by Takeuchi and colleagues in Brain Imaging and Behavior gave young adults four weeks of intensive training. Compared with training groups that did not speak (listening to compressed speech, or an active control task in the second language), the groups who practised shadowing or reading aloud in a second language showed a tendency towards greater increases in digit-span scores and significantly greater improvements in reaction time on a working-memory task. The study measured working memory, not conversational skill, so it is evidence that these exercises engage the system they are meant to, not that they make you fluent.

Wikipedia's article on speech shadowing describes it as a technique to practise speech in language learning, with tasks of increasing difficulty. How to do it without drilling errors is covered in the shadowing technique explained.