Hashtagging photos on Instagram represents a recent iteration of user-generated content organization. This thesis examines whether established theories of classification can explain these hashtagging practices by asking the questions "Are early conceptions of user generated classifications useful descriptors of hashtagging practices on Instagram?" And "Do older classification theories, developed prior to hashtagging practices still apply in a user generated context?" The first examines nonformal classification systems such as folksonomies while the second examines Hacking's (1986, 1996) dynamic nominalism, Bowker and Stars' (1999) case study informed definitions, and perspectives on the role of ontologies in classification. A hybrid walkthrough methodology was applied in Instagram to empirically examine the classificatory processes of content producers for thirty-five sets of hashtag pools of photos. Overall most characteristics of formal classification systems apply to Instagram user generated content and unsurprisingly, hashtagging practices on Instagram are best characterized as folksonomic.