Justice STEVENS, with whom Justice MARSHALL joins, dissenting.
The Court's narrow construction of the word "seizure" represents a significant, and in my view, unfortunate, departure from prior case law construing the Fourth Amendment. Almost a quarter of a century ago, in two landmark cases-one broadening the protection of individual privacy, and the other broadening the powers of law enforcement officers -we rejected the method of Fourth Amendment analysis that today's majority endorses. In particular, the Court now adopts a definition of "seizure" that is unfaithful to a long line of Fourth Amendment cases. Even if the Court were defining seizure for the first time, which it is not, the definition that it chooses today is profoundly unwise. In its decision, the Court assumes, without acknowledging, that a police officer may now fire his weapon at an innocent citizen and not implicate the Fourth Amendment-as long as he misses his target.
For the purposes of decision, the following propositions are not in dispute. First, when Officer Pertoso began his pursuit of respondent, the officer did not have a lawful basis for either stopping or arresting respondent. See App. 138-140; ante, at 623, n. 1. Second, the officer's chase amounted to a "show of authority" as soon as respondent saw the officer nearly upon him. See ante, at 625-626, 629. Third, the act of discarding the rock of cocaine was the direct consequence of the show of authority. See Pet. for Cert. 48-49, 52. Fourth, as the Court correctly demonstrates, no common-law arrest occurred until the officer tackled respondent. See ante, at 624-625. Thus, the Court is quite right in concluding that the abandonment of the rock was not the fruit of a common-law arrest.
It is equally clear, however, that if the officer had succeeded in touching respondent before he dropped the rock-evenif he did not subdue him-an arrest would have occurred. See ante, at 624-625,626. In that event (assuming the touching precipitated the abandonment), the evidence would have been the fruit of an unlawful common-law arrest. The distinction between the actual case and the hypothetical case is the same as the distinction between the common-law torts of assault and battery-a touching converts the former into the latter. Although the distinction between assault and battery was important for pleading purposes, see 2 J. Chitty, Pleading *372-*376, the distinction should not take on constitutional dimensions. The Court mistakenly allows this common-law distinction to define its interpretation of the Fourth Amendment.
At the same time, the Court fails to recognize the existence of another, more telling, common-law distinction-the distinction between an arrest and an attempted arrest. As the Court teaches us, the distinction between battery and assault was critical to a correct understanding of the common law of arrest. See ante, at 626 ("An arrest requires either physical force . . . or, where that is absent, submission to the assertion of authority"). However, the facts of this case do not describe an actual arrest, but rather an unlawful attempt to take a presumptively innocent person into custody. Such an attempt was unlawful at common law. Thus, if the Court wants to define the scope of the Fourth Amendment based on the common law, it should look, not to the common law of arrest, but to the common law of attempted arrest, according to the facts of this case.
The first question, then, is whether the common law should define the scope of the outer boundaries of the constitutional protection against unreasonable seizures. Even if, contrary to settled precedent, traditional common-law analysis were controlling, it would still be necessary to decide whether the unlawful attempt to make an arrest should be considered a seizure within the meaning of the Fourth Amendment, and whether the exclusionary rule should apply to unlawful attempts.