# Machine Translation
**Domain:** Natural Language Processing / Artificial Intelligence
**Doc Type:** Discipline Node
**Maturity:** Developed
**Related:** [[Natural Language Processing]], [[IBM 701]], [[IBM Research]], [[Language Model]]
## Definition
**Machine translation** is the computational transformation of text or speech from one natural language into another. Its history moves through dictionaries and hand-built rules, statistical models, neural sequence systems, and large language models.
## IBM Lineage
The 1954 Georgetown–IBM demonstration used an [[IBM 701]] to translate a deliberately constrained set of Russian sentences into English. It was a public proof of possibility, not a general solution to translation. IBM continued work on translation across later decades, including research associated with the [[IBM Thomas J. Watson Research Center]].
## Ontological Importance
Translation reveals that language is neither a simple code substitution nor an entirely inaccessible human mystery. Meaning depends on syntax, context, culture, reference, and purpose. The field therefore sits at the intersection of [[Natural Language Processing]], [[Symbolic Language Engine]], representation, and [[Ambiguity]].
## Sources / Provenance
- IBM, “Machine-aided translation”: https://www.ibm.com/history/machine-aided-translation