The most common error in reading a patent is treating the abstract as if it were the invention. It is not. Under U.S. law, the scope of a patent is set by its claims, and the claims come in two basic forms — independent and dependent. An independent claim stands alone and recites every element needed to define the invention; it is the broadest statement of what the patent protects. A dependent claim refers back to another claim, pulls in all of that claim's limitations, and then adds at least one more, which makes it necessarily narrower.

This structure is fixed by statute. 35 U.S.C. 112(c) provides that "a claim may be written in independent or, if the nature of the case admits, in dependent or multiple dependent form." Section 112(d) then defines exactly how a dependent claim is read:

"A claim in dependent form shall contain a reference to a claim previously set forth and then specify a further limitation of the subject matter claimed. A claim in dependent form shall be construed to incorporate by reference all the limitations of the claim to which it refers."— 35 U.S.C. 112(d), source

Two consequences follow directly from that text. First, a dependent claim can never be broader than the claim it depends on, because it carries every one of that claim's limitations and then adds another. Second, to infringe a dependent claim, an accused product must meet the independent claim and the added limitation — so dependent claims are narrower targets but can survive even if the independent claim is later found invalid.

Reading the structure in a real patent

Tesla's granted US 8,875,828 B2, "Vehicle battery pack thermal barrier," shows the structure cleanly. Independent claim 1 recites a vehicle battery pack integrated assembly: a pack enclosure mounted under an electric vehicle between the cabin floor and the driving surface, plus a "multi-layer thermal barrier" interposed between the enclosure and the cabin floor that provides noise isolation, thermal isolation, and vibration damping, and that is built from a compressible elastic first layer and a thermally insulating second layer. Every element a competitor must have to fall within claim 1 is listed there. Nothing outside those elements is required.

The dependent claims then add quantified limitations. Claim 2 narrows claim 1 to a barrier "wherein said multi-layer thermal barrier has an acoustic insertion loss of greater than 10 dB for frequencies between 1000 Hz and 10 kHz," and claim 3 narrows it further to "greater than 20 dB" over the same band. Claim 6 adds a thermal-conductivity ceiling; claim 11 adds a compression-modulus floor. Each dependent claim incorporates all of claim 1's limitations and bolts on one measurable constraint. A product could practice claim 1 yet avoid claim 3 simply by having an insertion loss below 20 dB — which is precisely why the independent claim is the one that defines the outer boundary of protection.

Why the distinction is load-bearing

For anyone analyzing automotive IP, the independent-versus-dependent distinction is the difference between reading scope and reading detail. The abstract and the summary describe the invention in prose; they bind no one. The independent claims are the fence line. The dependent claims are fallback positions and engineering specifics — often the place where the real, defensible numbers live, but always subordinate in scope to the claim they reference.

The practical reading rule that follows from 35 U.S.C. 112 is simple: to know the broadest thing a patent covers, find the independent claims and read them limitation by limitation; to know the narrower fallback positions, read the dependent claims that hang off them. The abstract tells you the topic. Only the claims tell you the boundary.

Why patents carry both broad and narrow claims

A well-drafted patent almost never relies on a single claim, and the reason is risk management against the prior art. The independent claim reaches for the broadest defensible scope; the dependent claims build a ladder of progressively narrower positions beneath it. If a challenger later finds prior art that invalidates the broad independent claim, the narrower dependent claims — each carrying an additional limitation the prior art may not disclose — can survive. In the Tesla thermal-barrier patent, claim 1's general "multi-layer thermal barrier" is the broad reach, while claim 3's "greater than 20 dB" acoustic-insertion-loss requirement is a specific, measurable fallback. A piece of prior art might describe a generic two-layer barrier yet say nothing about a 20 dB insertion loss, leaving the dependent claim standing even if the independent claim falls.

This laddering also shapes how infringement is read. Because 35 U.S.C. 112(d) construes a dependent claim to incorporate every limitation of the claim it references, an accused product must satisfy the entire independent claim before any dependent claim can even be reached. Infringement of claim 3 of the Tesla patent therefore requires meeting all of claim 1 and the 20 dB limitation. Dependent claims narrow the target; they never widen it. That is the structural guarantee the statute builds in.

Multiple dependent claims and the abstract trap

Section 112 also recognizes a third form. A multiple dependent claim refers, in the alternative, to more than one earlier claim — a drafting convenience that the statute permits but constrains, since such a claim is read as incorporating the limitations of whichever single claim is being considered. The takeaway for a reader is the same: every dependent form, single or multiple, is narrower than what it references. None of this scope-setting work is done by the abstract or the written-description summary. Those passages exist to describe and contextualize the invention, and courts do not let them expand or contract the claims. The recurring mistake — quoting a patent's abstract as if it stated what the patent protects — collapses the moment you apply the statute: the specification concludes with claims that particularly point out and distinctly claim the invention, and it is those claims, read independent-first, that set the legal boundary. The statutory text is the authoritative source for how each claim form is construed.