Learn English – Data pronunciation: “dayta” or “dahta”

The prefix comes from a plural form of the Latin quantifier multus 'many'. As with other Latin plurals ending in -ī, the English pronunciations vary.

The Latin /ī/ was pronounced in Middle English as long /i:/ (as in Modern English seen), but the Great Vowel Shift (GVS), which turned Middle into Modern English, moved all the ME long vowels up a step on the vowel chart.

That meant that words like mice and house (/mi:s/ and /hu:s/ in ME), which were already at the top of the chart, couldn't go any further. So, to make room for ME /e:/ and /o:/ (as in seen and soon) at the top, they fell off and became diphthongs. In particular, ME high front /i:/ became ModE /ai/ , while ME high back /u:/ became ModE /au/.

Which is why mice and house are now pronounced /mais/ and /haus/. This means that multi- can be pronounced as in Latin or Middle English as /ˈməlti-/, or in post-GVS fashion as /ˈməltai-/.

The answer is all of the above, any of the above, or none of the above. There is no single set of rules for the pronunciation of taxa, and no single interpretation of such rules as some have attempted to compile. As Michael G. Simpson notes under "Pronunciation of Names" in Plant Systematics (2006),

Although scientific names are universal, their pronunciations may vary from region to region, especially between different countries. For example, European pronunciations are often different from those of most American botanists. There are no firm rules as to how scientific names should be pronounced. Very often, pronunciations are influenced by one's native language. One should be flexible and adaptive with regard to pronunciations, as the overriding goal is communications.

Broadly, most English-speaking practitioners of science, medicine, law, architecture, and other disciplines follow traditional English pronunciation of Latin for the Latin and quasi-Latin terms in their fields. There are those who push for what is known as Reformed Academic Pronunciation, a system devised in the late 19th century which is supposed to be closer to the classical pronunciation, and which is closer to the way the words would be pronounced in most continental European languages. William Stearn favors the latter in his widely cited Botanical Latin (1983), but as countless papers, guides, and appendices note:

• Professionals not only use different pronunciations from one place to another, but do not consistently follow the same system themselves.
• People tend to pronounce names based on how they first hear them, as opposed to a particular system of pronunciation
• English speakers don't agree on how to pronounce English; how would they agree on how to pronounce any other language or pseudo-language? (Yes, I am one of those people who is always going on about bruschetta— it's Italian, not French or German.)

Zoologists and bacteriologists moved away from requiring a classical basis for names some time ago, and botanists have moved in the same direction, so perhaps there will be less debate in the future.