Skip to content
PatentBrief

Patent Drafting

Claim Language Guide

Every word in a patent claim carries legal weight. "Comprising" opens the door to additional elements; "consisting of" slams it shut. The preamble may or may not limit. "Means for" shrinks scope to what was disclosed.

FAQ

What do 'comprising,' 'consisting of,' and 'consisting essentially of' mean in patent claims?

The transitional phrase between the preamble and the body of a claim is one of the most important words in patent law: COMPRISING (open transition): the most common transitional phrase; means the claim is OPEN — the claim is satisfied even if the accused product or process contains additional elements not listed in the claim; example: 'A composition comprising A, B, and C' covers a composition containing A, B, C, and also D — because D is an unlisted addition, not a substitution; COMPRISING = 'includes at least' or 'having at least'; CONSISTING OF (closed transition): the claim is CLOSED — the claim is satisfied ONLY if the accused product contains exactly the listed elements and nothing else; example: 'A composition consisting of A, B, and C' does NOT cover a composition containing A, B, C, and D — the presence of D defeats infringement; CONSISTING OF is used when it is important to define a precise composition and exclude any other components; CONSISTING ESSENTIALLY OF (semi-open transition): the middle ground; the claim is open to additional elements UNLESS those elements materially affect the basic and novel properties of the claimed invention; PPG Industries v. Guardian Industries (Fed. Cir. 1997): the 'basic and novel properties' are determined by the specification and prosecution history; used frequently in chemistry and materials science claims; WHICH TO CHOOSE: 'comprising' provides the broadest protection and is the default; 'consisting of' is used when the composition must be precisely defined (e.g., pharmaceutical formulations, alloys); 'consisting essentially of' is intermediate and is litigated over the meaning of 'materially affect'; NESTED CLAIMS: a dependent claim can use 'consisting of' to close an open 'comprising' independent claim — this allows prosecution with a broad comprising independent claim and a narrow consisting-of dependent claim as a fallback.

When does the preamble of a patent claim limit its scope?

The preamble is the introductory phrase before the body of a claim — and whether it limits claim scope is a notoriously unsettled area of patent law: WHAT IS THE PREAMBLE: in 'A widget comprising: A, B, and C' — the preamble is 'A widget'; in 'A method of treating a disease comprising: administering A' — the preamble is 'A method of treating a disease'; GENERAL RULE: preambles generally do NOT limit claim scope — they are merely a recitation of the intended purpose or field; the claim elements in the body (the 'comprising' list) do the limiting work; WHEN THE PREAMBLE DOES LIMIT (exceptions): (a) ANTECEDENT BASIS: if an element in the body of the claim uses the preamble term as its antecedent basis, the preamble is limiting; example: preamble says 'a computing device'; body says 'the computing device further comprising' — 'the computing device' refers back to the preamble, making it limiting; (b) LIFE OF THE CLAIM: if the preamble is essential to understanding the body of the claim, it limits — the body cannot make sense without it; (c) INVENTOR'S STATEMENT: if during prosecution the inventor argued that the preamble distinguished prior art, the preamble may limit; (d) SPECIFICATION AND PROSECUTION HISTORY: if the specification or prosecution history treats the preamble as a limitation; PRACTICAL IMPACT: a limiting preamble adds an additional element that must be present for infringement; a non-limiting preamble can be ignored in infringement analysis; STRATEGY: if the preamble element is easy to find in accused products, making it non-limiting is preferred; if the preamble distinguishes prior art (provides patentability), it is likely limiting.

What is means-plus-function claiming under § 112(f) and when does it apply?

Section 112(f) allows functional claiming limited by the disclosure in the specification — a narrower, disclosure-dependent scope: STATUTORY BASIS: 35 U.S.C. § 112(f): 'An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof'; HOW MPF WORKS: instead of claiming 'a bolt and nut fastener,' an MPF claim says 'means for fastening'; scope is NOT the full universe of fasteners — only the specific structure disclosed in the specification plus 'equivalents'; TRIGGERS: 'means for [function]' or 'step for [function]' language typically invokes § 112(f); Williamson v. Citrix Online (Fed. Cir. 2015 en banc): 'nonce words' (module, unit, component, element, mechanism, system, device) that perform a function WITHOUT sufficient structure can also invoke § 112(f); CONSEQUENCE: MPF claims are narrower than they appear — limited to disclosed structure; if the specification discloses only one algorithm for a software function, the claim covers only that algorithm and structural equivalents; INVALIDITY RISK: if the specification does NOT disclose structure corresponding to the claimed function, the claim is indefinite (§ 112(b) invalidity); Aristocrat Techs v. IGT (Fed. Cir. 2007): software MPF claims must disclose the specific algorithm, not just a general-purpose computer; AVOIDING MPF: if the patentee wants broad functional claiming without the disclosure limitation, they should include structural recitation (e.g., 'a processor configured to...'); this avoids invoking § 112(f) while still reciting function; MODERN TREND: patent drafters generally avoid 'means for' language to prevent the narrow MPF interpretation, instead using structural language that is functionally described.

What are key rules for interpreting claim element language (articles, numerals, and ranges)?

Patent claim elements use articles and quantifiers with precise legal meanings: ARTICLES — 'A' and 'AN': introduce an element; mean 'one or more' in comprising claims — not 'exactly one'; 'a bolt' in a comprising claim covers a fastener with multiple bolts; Baldwin Graphic Systems v. Siebert (Fed. Cir. 2008): 'a' means one or more; THE DEFINITE ARTICLE: 'the' or 'said' — requires antecedent basis; 'the bolt' must refer back to 'a bolt' introduced earlier in the claim; SINGULAR vs. PLURAL: if the claim says 'a plurality of bolts,' it requires more than one; if it says 'a bolt,' it means one or more; 'first' and 'second': ordinal language ('first widget,' 'second widget') typically means the widgets are distinct — two different widgets, not the same one; NUMERICAL LIMITATIONS: 'at least one' = one or more; 'exactly one' or 'one and only one' = exactly one; 'one or more' = same as 'at least one'; 'two or more' = at least two; without explicit numerical limitation, 'a' means one or more; RANGE CLAIMS: 'between X and Y' is ambiguous (does it include endpoints?); patent claims should use 'from X to Y' with 'inclusive' or 'exclusive' specified; court construes 'between' based on specification and claim context; RANGES IN DIFFERENT ELEMENTS: Dow Chemical v. Astro-Valcour: when a claim recites two ranges that partially overlap prior art, the entire claim may still be patentable if the intersection is novel; NEGATIVE LIMITATIONS: 'wherein the composition is free of alcohol' is a negative limitation; valid if supported by the specification; Inphi Corp. v. Netlist (Fed. Cir. 2014): negative limitations are permissible and limit scope.

What are common claim drafting mistakes and how do they affect patent protection?

Claim drafting errors can significantly narrow or invalidate patent protection: MISTAKE 1 — ANTECEDENT BASIS ERRORS: introducing 'the widget' without first introducing 'a widget'; this creates indefiniteness under § 112(b) — what does 'the widget' refer to?; fix: ensure every definite article reference has a prior introduction; MISTAKE 2 — IMPORTING LIMITATIONS FROM EXAMPLES: a claim that uses language from the specification's preferred embodiment may be construed to be limited to that embodiment; Honeywell v. ITT (Fed. Cir. 2004): 'fuel injection system component' construed as limited to a fuel filter because the specification described only fuel filters; fix: write claims with abstract functional language; disclose multiple embodiments in the specification; MISTAKE 3 — NARROWING IN PROSECUTION WITHOUT RESERVING BROADER CLAIMS: when adding a limitation to overcome a prior art rejection, failing to file a continuation with the original broader claim; fix: file a continuation before allowance to preserve the broader claims in a sibling application; MISTAKE 4 — OVERUSING 'MEANS FOR' LANGUAGE: triggers § 112(f) MPF interpretation, limiting scope to disclosed structure; fix: use 'a processor configured to perform X' rather than 'means for performing X'; MISTAKE 5 — CLAIMING THE PREFERRED EMBODIMENT ONLY: drafting claims that cover only the currently manufactured product, leaving competitors free to modify one element; fix: draft independent claims at the highest level of abstraction the prior art permits; MISTAKE 6 — FAILING TO CLAIM METHOD AND SYSTEM SEPARATELY: a method claim and a system claim for the same invention have different infringement standards (35 U.S.C. § 271(a) for direct infringement; § 271(b) for induced); draft both method and apparatus claims to cover different infringement scenarios; MISTAKE 7 — FORGETTING DEPENDENT CLAIMS: independent claims may be invalidated; dependent claims provide fallback protection; draft narrow dependent claims capturing commercially important specific embodiments.

Related Guides

Claim ScopeClaim ConstructionDependent ClaimsMeans-Plus-FunctionProsecution History EstoppelPortfolio Strategy