Lilly does not contend that consumers are likely to believe that HERBROZAC capsules are really PROZAC ®. Such confusion is not likely because one cannot even purchase PROZAC ® without a prescription from a licensed physician, who is not at all likely to confuse one product for another. Lilly contends, however, that the similarity of the trademarks is likely to confuse some consumers as to whether there is some affiliation or association between the sources of the two products. The Lanham Act applies to use of a product name that "is likely to cause confusion, or to cause mistake, or to deceive as to the affiliation, connection, or association of such person with another person, or as to the origin, sponsorship, or approval of his or her goods, services, or commercial activities by another person .
§ 1125(a) (1); see also 4 McCarthy, § 24:16 *843 at 24-33 ("If the consumer is reasonably mistaken as to the source or sponsorship of an alleged infringer's goods, she suffers a real and independent injury . Tour 18 I Ltd., 155 F.3d 526, 543 (5th Cir.1998) (golf course designed to imitate famous golf holes found likely to cause confusion of affiliation between the original golf courses and imitation course). On the issue of consumer care with respect to affiliation or association, there is no reason to expect consumers of HERBROZAC to exercise great care to determine whether or not the seller of HERBROZAC is affiliated, connected, or associated with the seller of PROZAC ®. "The term `strength' as applied to trademarks refers to the distinctiveness of the mark, or more precisely, its tendency to identify the goods sold under the mark as emanating from a particular . Quaker Oats Co., 978 F.2d 947, 959 (7th Cir.1992), quoting McGregor-Doniger Inc. Drizzle Inc., 599 F.2d 1126, 1131 (2d Cir.1979), superseded by rule on other grounds, as recognized in Bristol-Myers Squibb Co. McNeil-P.P.C., Inc., 973 F.2d 1033, 1044 (2d Cir.1992). First, it is a fanciful word that carries no meaning apart from its use to identify a product.
It does not describe or suggest the function of the product it names. When the seller of a product coins a word just for the product, as Lilly did with PROZAC, trademark protection is at its highest. Polaraid, Inc., 319 F.2d 830, 837 (7th Cir.1963); accord, Hyatt Corp. Hyatt Legal Services, 736 F.2d 1153, 1158 (7th Cir.1984). Natural Answers asserts that PROZAC ® is now so famous that it has become a "generic" term no longer entitled to trademark protection. A trademark becomes generic when it comes to describe, in common usage in the United States, a class of goods rather than an individual product. New America Publishing, Inc., 971 F.2d 302, 306 (9th Cir.1992). The Seventh Circuit has explained: When a trademark becomes generic, such as "aspirin" or "thermos," and so loses trademark protection, because the public, perhaps egged on by the omni-present media, decides to use the trademark to designate not the particular manufacturer's brand but the entire product comprising all the competing brands, the trademark is dead no matter how vigorously the holder has tried to prevent this usage. GTE Vantage, Inc., 99 F.3d 244, 247 (7th Cir.1996) (collecting cases and affirming finding that "March Madness" phrase was generic). Where the plaintiff's trademark has been registered with the United States Patent and Trademark Office, the plaintiff is entitled to a presumption that the mark is not generic, and the burden of proof is on the defendant to come forward with evidence showing that the mark is generic. Liquid Control Corp., 802 F.2d 934, 936 (7th Cir.1986); accord, Reese Publishing Co. Hampton Int'l Communications, Inc., 620 F.2d 7, 11 (2d Cir.1980), citing McGregor-Doniger Inc. Drizzle, Inc., 599 F.2d at 1132; Venetianaire Corp. For an illustration of the type of proof useful in showing that a trademark has become descriptive and generic, the district court's opinion in the "thermos" case is instructive. The evidence in the thermos case included hundreds of newspaper and magazine articles, as well as dictionary and encyclopedia references, large volumes of *844 correspondence, and a public survey. The district court found that the large volume of references using the term "thermos" in a generic or descriptive sense showed that the term had lost its trademark status as identifying the source of a particular brand of vacuum-insulated containers. The Second Circuit affirmed on the district court's detailed findings. Professor McCarthy has catalogued types of evidence useful in proving that a term has become generic. Such evidence may include generic use of the mark by competitors without objection from the plaintiff, the plaintiff's own use in a generic way, dictionary definitions, generic use in the media, testimony of persons in the trade about the use of the term, and consumer surveys. Natural Answers has tried to use Lilly's own evidence of the fame of PROZAC ® to show that the term has become generic. Natural Answers highlights, for example, the Newsweek statement that "Prozac has the familiarity of Kleenex and the social status of spring water." See, e.g., Pl.Ex. In fact, however, the cluster of Newsweek articles did not use the term generically but distinguished between PROZAC ® and other antidepressant drugs. A trademark can become exceptionally strong and famous without becoming generic, as suggested by the examples of COCA-COLA ® and KODAK ®. Natural Answers has not shown that PROZAC ® is likely to be deemed generic.
There is no evidence that any other competitors have used the term in a generic way, let alone that Lilly has tolerated such use.
 There is no evidence that Lilly itself has used the term in a generic way. The parties have not submitted dictionary definitions tending to show the term has seeped into common usage in a generic sense. See Illinois High School Ass'n, 99 F.3d at 246 (trademark holder must try to convince "dictionary editors, magazine and newspaper editors, journalists and columnists, judges, and other lexicographically influential persons" to avoid using the trademark in a generic way). The media references actually before the court do not use the term "Prozac" in a generic way. They use the term instead to identify plaintiff Lilly's product.