dissenting.
This case presents a conundrum often faced by district courts when construing *989claims: Whether and how to consider claim-limiting language found only in the specification. Our court has not been silent on this issue:
We have had many occasions to cite one or both of the twin axioms regarding the role of the specification in claim construction: On the one hand, claims “must be read in view of the specification, of which they are a part.” Markman v. Westview Instruments, Inc., 52 F.3d 967, 979 (Fed.Cir.1995), aff'd, 517 U.S. 370, 116 S.Ct. 1384, 134 L.Ed.2d 577 (1996). On the other hand, it is improper to read a limitation from the specification into the claims. Arlington Indus., Inc. v. Bridgeport Fittings, Inc., 345 F.3d 1318, 1327 (Fed.Cir.2003); Gart v. Logitech, Inc., 254 F.3d 1334, 1343 (Fed.Cir.2001). Although parties frequently cite one or the other of these axioms to us as if the axiom were sufficient, standing alone, to resolve the claim construction issues we are called upon to decide, the axioms themselves seldom provide an answer, but instead merely frame the question to be resolved.
Liebel-Flarsheim Co. v. Medrad, Inc., 358 F.3d 898, 904 (Fed.Cir.2004).
I fear the majority, armed with too rigid an interpretation of the court’s recent decision in Phillips v. AWH Corp., 415 F.3d 1303 (Fed.Cir.2005) (en banc), has not heeded its own admonition and now overturns the district court’s construction of the term “plug” based on the latter axiom standing alone.
It is good policy, of course, to require patent writers to place all claim limitations in the claims of the patent. Section 112 of the Patent Act gives clear marching orders in this regard: “The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.” 35 USC § 112 ¶ 2. The claims define the inventor’s right to exclude, and, just as a patent holder should not be allowed to broaden an inventor’s claims based on the written description, construing courts .must take care not to narrow claims unduly.
But patents must be read as well as written. Judges, parties and the public must understand claim terms by giving them their “ordinary and customary meaning.” Phillips, 415 F.3d at 1312 (internal quotation omitted). The ordinary and customary meaning of a claim term is “the meaning that the term would have to a person of ordinary skill in the art in question at the time of the invention * * Id. at 1313. This person of ordinary skill in the art “is deemed to read the claim term not only in the context of the particular claim in which the disputed term appears, but in the context of the entire patent, including the specification.” Id.
The majority concludes that “‘[p]kig’ should * * * be given its ordinary meaning: ‘a member movable between an open and closed position,’ ” without ever explaining why it considers this to be the “ordinary” meaning. Perhaps the majority considers this to be the ordinary meaning because it is similar to the broad dictionary definitions offered by Medegen: “a piece of wood, metal, or other material used or serving to fill a hole” or “a stopper, stopple or bung.” But construing claim terms based on dictionary definitions or even on the court’s existing understanding of commonly used words — the “a plug is a plug” line of reasoning — risks polluting the claim construction analysis with the biases the Phillips court warned of when observing that “extrinsic evidence in general, [is] less reliable than the patent and its prosecution history in . determining how to read claim terms * * *.” Phillips, 415 F.3d at 1318-19.
*990The very able district court, on the other hand, looked first to the patent in its effort to construe “plug” as a person of ordinary skill in the art of medical connectors would after reading the entire '418 patent. In its claim construction order, the district court observed that the specification and the figures of the '418 patent repeatedly show the plug to have the elastomeric, pivoting and buckling limitations the district court included in its construction of ‘plug.’ The district court was cognizant of our warnings against confining claims to a preferred embodiment, but, after examining the entire patent, determined that the claimed invention would not work without the plug having the elastomeric and buckling or pivoting limitations. The specification’s consistent descriptions of the plug as elastomeric and pivoting or buckling disclosed an essential element of the invention — those characteristics prevent the syringe from forming a seal with the plug and allow the opening of the fluid path necessary for the invention to work. See, e.g., '418 patent fig 2. Having determined that the specification suggests a narrower definition of “plug” than that advocated by Medegen, the court then turned to extrinsic evidence: it asked Medegen’s counsel how a rigid plug would work in Medegen’s device, and did not receive a satisfactory response. The expert report submitted by ICU also supports the court’s finding that one skilled in the art of medical connectors having read the '418 patent at the time of invention would have understood the “plug” to be elastomeric and able to pivot and buckle.
The majority relies heavily on Phillips’s rejection of the “contention that if a patent describes only a single embodiment, the claims of the patent must be construed as being limited to that embodiment.” Phillips, 415 F.3d at 1323. But the district court here did not determine that the patent must be construed as limited to a single embodiment simply because just one embodiment was described. Rather, it determined that the invention described in the '418 patent would not work without the limitations to “plug” described in the specification. This was not the case in Phillips, where our court noted that some of the embodiments in the patent at issue did not include the limitations improperly incorporated into the claim construction by the district court. Id. at 1325-26 (“Importantly, Figures 4 and 6 do not show the baffles as part of an ‘intermediate interlocking, but not solid, internal barrier.”). The claim limitations at issue in Phillips were not essential to the invention; here, it appears, they are.
“[I]n a given case, the scope of the right to exclude may be limited by a narrow disclosure.” Gentry Gallery v. Berkline Corp., 134 F.3d 1473, 1479 (Fed.Cir.1998). It appears that this may be such a case. As a jurist more accustomed to working on the front lines of patent litigation than reviewing decisions from above, it is my experience that claim construction — determining how one of ordinary skill in the art would understand the patent at the time of invention' — often requires making fact-like determinations not well suited to appellate review.
My concern is that the majority has exaggerated the utility of the axiom that a limitation from the specification should not limit the claims. This results in sending the case back to the district court almost two-and-a-half years after it commenced, and almost a year and a half after the Markman hearing, for the preliminary step of construing the claim terms. While guidance from this court is essential to creating uniformity and predictability in patent litigation, the district court here did not depart from our basic instructions. As Judge Lourie wrote of the claim construction in Phillips:
*991[E]ven though claim construction is a question of law, reviewable by this court without formal deference, I do believe that we ought to lean toward affirmance of a claim construction in the absence of a strong conviction of error. I do not have such a conviction in this case, after considering the district court’s opinion and the patent specification.
Phillips, 415 F.3d at 1330 (Lourie, J, concurring in part and dissenting in part).
I, therefore, respectfully dissent.