Definition:
Surface structure (also S-structure in Government and Binding Theory) is the level of syntactic representation that directly corresponds to the phonological form of a sentence — the actual words in their actual order as spoken or written. It is the output of the transformational rules that operate on the underlying deep structure. While deep structure encodes abstract semantic relations, surface structure is what speakers produce and what listeners hear.
The Surface/Deep Contrast
In Chomsky’s transformational grammar, a sentence passes through (at least) two syntactic levels:
Deep structure: The abstract underlying form encoding basic semantic relationships.
Surface structure: The phonologically realized form after transformation.
Example — Wh-question:
Deep structure:
> John saw what? (the questioned element what is in its canonical object position)
After Wh-movement transformation (move what to sentence-initial position + invert auxiliary):
Surface structure:
> What did John see ___?
The position marked ___ (a “trace” in GB Theory) shows where what originated — connecting surface and deep structure.
Example — Passive:
Deep structure:
> [NP the police] arrested [NP the rioters]
After passive transformation:
Surface structure:
> [NP The rioters] were arrested [by the police]
The surface order is reversed, but the underlying agent-patient relations are preserved at deep structure (and recoverable by interpretation rules).
What Surface Structure Feeds
In the Standard Theory (Chomsky 1965), surface structure is the input to:
- Phonological rules — which handle pronunciation, stress, and intonation
- Certain semantic interpretation rules (scope of quantifiers, topic-focus structure)
In later frameworks (Government and Binding), S-structure also feeds:
- The phonological component (PF — Phonological Form)
- Certain binding and Case-checking principles
In the Minimalist Program (1995 onward), the explicit S-structure level was eliminated — Merge operations feed directly into two interface levels (PF and LF — Logical Form), without an intermediate S-structure.
Surface Structure and Ambiguity
One function of positing deep vs. surface structure: explaining structural ambiguity at the surface.
“The shooting of the hunters was terrible.”
- Surface form: NP is ambiguous
- Deep structure 1: [someone shot the hunters] → the hunters were shot
- Deep structure 2: [the hunters were shooting] → the hunters’ shooting (was bad)
Both readings have the same surface structure but distinct deep representations.
Surface Structure and SLA
Surface vs. underlying knowledge is a recurring issue in SLA:
- Learners sometimes produce correct surface forms for the wrong reasons (they memorized a chunk but don’t have the underlying rule)
- Learners sometimes produce incorrect surface forms despite having correct underlying representations (their rule is right but they haven’t automatized the surface realization)
- Error analysis focuses on surface output, but understanding errors requires hypothesizing about underlying representations
The gap between learner competence (deep grammatical knowledge) and performance (surface output) is a persistent methodological challenge in SLA research.
History
Surface structure was introduced by Noam Chomsky in Syntactic Structures (1957) and elaborated in Aspects of the Theory of Syntax (1965) as part of transformational grammar. In this framework, sentences have two levels of representation: deep structure (the underlying abstract representation of meaning and grammatical relations) and surface structure (the actual linear order of words as spoken or written). Transformational rules map deep structure to surface structure — for example, the passive transformation converts “The cat chased the mouse” (active surface structure) from the same deep structure that generates “The mouse was chased by the cat” (passive surface structure). The deep/surface distinction was central to early generative grammar but was significantly revised in later frameworks: Government and Binding Theory (1981) reinterpreted the levels, and the Minimalist Program (1995) largely eliminated the distinction in favor of a single derivational process.
Common Misconceptions
“Surface structure is what a sentence ‘really means.’”
Surface structure is the form — the linear sequence of words. Meaning is associated with the deep/underlying structure. The insight of transformational grammar is that different surface forms can share the same meaning (paraphrases) and identical surface forms can have different meanings (ambiguity).
“Surface structure is unimportant because deep structure carries the meaning.”
Surface structure is what listeners actually hear and process. Understanding how surface structure relates to underlying meaning is essential for parsing, comprehension, and production. Second language learners must learn the surface structure patterns of the target language to communicate.
“Modern linguistics still uses the deep/surface structure distinction.”
The specific deep/surface distinction of 1960s transformational grammar has been largely replaced. Modern generative grammar uses different representational levels, and many non-generative frameworks reject the distinction entirely. The concepts remain useful for introductory linguistics but do not represent current theoretical consensus.
“Surface structure is the same across all languages expressing the same meaning.”
Different languages use dramatically different surface structures to express similar meanings. Japanese SOV order (猫がネズミを追いかけた) and English SVO order (“The cat chased the mouse”) are different surface structures for similar semantic content.
Criticisms
The surface structure concept, as part of the deep/surface distinction, has been criticized from multiple perspectives. Usage-based and construction grammar approaches reject the need for two levels of syntactic representation, arguing that grammatical knowledge consists of learned form-meaning pairings (constructions) at a single level without abstract deep structures and transformations.
Even within generative grammar, the deep/surface distinction in its original form was abandoned: the Minimalist Program reconceives syntactic derivation as a single computation that generates structure through successive applications of Merge, without distinct “deep” and “surface” levels. The original distinction also faced empirical challenges — many syntactic phenomena could not be cleanly explained as transformations from deep to surface structure, requiring increasingly complex rule systems that undermined the explanatory elegance the theory intended.
Social Media Sentiment
Surface structure rarely appears in language learning communities. It is encountered in introductory linguistics courses and occasionally referenced in discussions about why translations between languages sound unnatural — learners sometimes note that “you can’t just rearrange the surface structure of one language into another.”
The concept is more culturally familiar than its technical content suggests: the everyday distinction between “what someone said” (surface) and “what they meant” (deeper meaning) echoes the spirit of the deep/surface distinction, though this popular understanding diverges significantly from the technical linguistic concept.
Practical Application
- Learn the surface structure patterns of your target language — Word order, particle placement, and morphological marking differ across languages. Japanese surface structure (SOV with particles) requires different processing strategies than English (SVO with prepositions).
- Don’t translate surface structures literally — Attempting to map the word order of your L1 onto your L2 produces ungrammatical output. Learn the target language’s patterns on their own terms.
- Recognize that different surfaces can express the same content — Active/passive alternations, topic-comment vs. subject-predicate structures, and other syntactic variations create different surface forms for similar underlying content. Exposure to these variations improves comprehension flexibility.
- Practice both recognition and production — Understanding surface structure patterns in input (reading, listening) and producing them in output (speaking, writing) require different skills. Both need deliberate practice.
Related Terms
See Also
Research
Chomsky (1957, 1965) established the deep/surface structure framework. The subsequent history of the distinction — its revision in Government and Binding Theory (Chomsky, 1981) and its replacement in the Minimalist Program (Chomsky, 1995) — reflects the evolution of generative syntactic theory.
For SLA, the deep/surface distinction informed early research on error analysis: surface-level errors (word order, morphological marking) were distinguished from deep-level errors (underlying structural misanalysis). Pienemann’s (1998) Processability Theory proposes that L2 learners acquire surface structure operations in a predictable sequence based on their processing demands — simpler surface operations (canonical word order) before more complex ones (movement and subordination) — providing a developmental framework that connects formal syntax to acquisition research.