七 名之構成
Seven Components of the Name
如吾人噭然而曰「犬」，用茲「犬」之一名，名一現觀之物，則亦以吾現觀之此一物，與吾曾知之犬相類，或其形色相類，或其質觸相類，或其吠聲相類，或其活動相類，或其表示於聲動之心態相類，吾乃將此所現觀之一物，納入吾憶念中所曾知之「犬共相」中──一共相理，即一概念──，名之曰犬。若察其但有形色之相類，則名犬畫犬影而不得名犬也。若察其但有形色質觸之相類，則名人造物犬而不得名犬也。雖察其亦有聲音活動而無表示之心態，則名人造機器犬而仍不得名犬也。此為「犬共相」內非真犬之分別。使其為真犬也，或疑曰「羊」，然又當憶念吾曾知之羊共相，辨其形色非羊，質觸非羊，聲音非羊，動態非羊，心態非羊，乃得決其疑曰「此必是犬而定非羊」。對於疑曰羊者如此，對於疑曰牛者、馬者、人者、樹者、石者、雲者，如是乃至歷窮一切名物，皆可如此。此為「犬共相」外非犬類之分別。盡此內容及外延之類別，對於此一物屬於犬共相乃得疑斷信成，名之曰犬。故加一物以「名」，其名必經比量乃獲構成。其式如下：
此一物類吾人曾知之犬………因…………小前提
既類吾人曾知之犬應是犬……同喻體……大前提
A dog, as we understand it, should be a dog... The same analogy applies... The major premise.
夫以一名指稱一物，習若甚易，乃必經許多之推比，始獲決知而下「此是犬」之結論──宗──，名之曰犬，竟如是艱難也！然此猶以「憶知中有約定俗成之共相名」為所依者。若使吾人新遇一物，於習俗中既無成名可以名之，吾曾知之名物中亦絕無其相類者，雖可刱製一名名之，然欲使此刱製之名盡能詮表此所名物共相內共相外之類別理，必察盡此物共相內之別以類其內，又辨盡此物共相外之類以別其外，極成為是此非餘之概念，確能以此名詮此概念──概念即物──，此名乃為成立。既為他日重遇此物同類之物可加之以此名，亦為曉喻他人可用此名以說明於此物。凡此或以曾知之類名名現觀之物，或觀新遇之物而刱新名，以理立名，以名詮理，名物──即名理──之構成皆經比量，決然可知之矣。名有遍名者，前所謂最大共，則都無不共也。名有獨名者，前所謂最小共，則但有不共也。名有類名者，為有不共之大共名──若言生物，對無生物為不共，對生物為共──。名有別名者，為有不共之小共名──若言人生，對旁生等為不共，對人等為共──。類上有類，則類名可退為別名；別下有別，則別名可進為類名。雖進退之無定，其為類名別名無異。剋實言之，唯有類名別名，無遍名與獨名。中國之「物」，似為遍名，其廣義無不可名為「物」故。然物仍與非物相對，非物即「無」，物之一名不遍於「無」，即非遍名，使為遍名，則吾人在分別記憶中即不能有「物名」之概念。佛學之「法」，似為遍名，有曰有法，無曰無法，雖無亦可名為法故。然法仍與非法相對，使為非法，吾人分別憶念中即不能有「法名」之概念；然今姑以法為遍名。在一「一神教」於所奉之「神」，似為獨名，然合其餘神教所奉之神，則亦為類下之別名，或別上之類名。在一世界之「日」，似為獨名，然日為恆星而類於日之恆星，太空中不知有若干千萬，故日亦為類中別名。一人「姓名」，似為獨名，而古今亦多同此姓名者，故亦可為類中別名。然今姑以若「孔丘」，若「釋迦牟尼」等為獨名。遍名但便總持，不彰比量之理；獨名但便稱呼，亦無比量之理。故比量之理智，著於類別之名。
Take the simple case of pointing to an object and calling it "dog." This seems easy, but in reality, it takes many rounds of reasoning and comparison before we can confidently conclude, "This is a dog"—the thesis—and name it as such. Even this is based on a conventional, shared concept of a "dog" already present in our memory.
Now imagine encountering a completely new object—one with no established name in our culture and no resemblance to anything we already know. We could invent a name for it. But for that new name to accurately express both the internal and external categories of the named object, we must thoroughly examine the internal distinctions to classify it within its type, and then differentiate it from all external types to make it uniquely "this and not that." Only when we have formed a precise concept—where concept equals object—can the name be properly established. This name would then be used to identify similar objects in the future and to communicate information about it to others.
Whether we use an existing class name to label a perceived object, or we observe a new thing and create a new name based on rational principles, the construction of a name-object (i.e., a name-concept) always involves inference and reasoning. This is certain.
Names can be universal—like the "supreme commonality" mentioned earlier, which has no exclusive aspects. Names can also be unique—like the "minimal commonality," which has only exclusive aspects. Then there are class names, which are common with some exclusivity: for example, "living being" is exclusive compared to "non-living thing," but common among living beings. There are also specific names, which are shared within a smaller category: "human" is exclusive compared to "other animals," but common among humans.
Class names and specific names are relative. A class name can become a specific name when placed above a higher class; a specific name can become a class name when placed below a subclass. Though their position may shift, their nature as class name or specific name remains the same. In truth, only class names and specific names exist; there are no truly universal or unique names.
In Chinese, "thing" (物) seems universal, as almost everything can be called "thing." But "thing" is still relative to "non-thing" (nothingness); "thing" does not encompass "nothing," so it is not truly universal. If it were, we could not conceptually grasp "thing" in our discriminating memory. In Buddhism, "dharma" (法) seems universal, as both existence (有法) and non-existence (無法) can be called "dharma." Yet "dharma" still stands in opposition to "non-dharma"; if something were "non-dharma," we would not have the concept of "dharma" in our discriminating memory. For now, we may treat "dharma" as a universal name.
In a monotheistic religion, the name of the worshipped "God" seems unique. But when compared with gods worshiped by other religions, it becomes a specific name within a class or a class name above a subclass. Our world's "Sun" seems unique, but the sun is a star, and there are countless such stars in the universe—so "sun" is really a specific name within a class. A person's "name," like "Confucius" or "Siddhartha Gautama," seems unique, but many people throughout history share the same name. So names like "Confucius" or "Siddhartha Gautama" are considered unique.
Universal names are convenient for summary but do not reveal the logic of inference. Unique names are convenient for reference but also lack inferential logic. Therefore, the rational understanding derived from inference is rooted in class and specific names.