The Warchief

2022 brought me an exciting new opportunity - Stronghold, who had employed me to work on one of their biggest fantasy projects, were so pleased with the quality of my work they invited me onto another one!

This was Warchief, an epic fantasy film that featured two excellent Orc antagonists: the conniving, manipulative Warshaman, and the fearsome Warchief himself. These were characters whose culture was defined by war, but we wanted to make sure that our approach was a refreshing one - instead of the traditional grunting, dim-witted Orcs, we wanted them to have a way to express themselves fluidly and intelligently while still preserving their intimidating nature. One scene revolves around the clash of communication between the Orcs and their human opponents, so it needed to be distinctly different.

Once again, it was my job to create a fully-fledged Orcish language, create PowerPoints each week for progress meetings, and translate every line of Orcish in the script with a phonetic guide for each, all before the shooting date - a very complex task indeed!

This, it turned out, would be my biggest linguistic challenge yet. My work on the other fantasy film had been a relatively smooth and comfortable foray into complex linguistics, but the Orcish language represented a massive step up in difficulty. This was spurred almost entirely by the non-human nature of the Orcs - I wanted to develop this language based on their traits from the ground up, taking into account their physiology as well as their culture. I was given a full overview of the workings of the other departments through Trello, as I was joining the project late into development, and immediately pinpointed a detail that interested me: the shape of the Orcs' mouths and their enormous prosthetic tusks.

A lot of language design depends on a strong foundation of phonemes - the sounds we make and string together to produce words. These, in turn, are defined by the placement and movement of the lips and tongue and the way we move air through those shapes. Compare, for example, "bot" and "boot". Although they are extremely similar, we know that the "o" sound is different from an "oo" sound, just like "boot" and "boat". Our accents - differences in the way we make mouth-shapes - will change how these are pronounced, and that was another challenge I had to conquer. The two lead actors had very different accents, so I had to make sure that any documentation I made would be able to articulate the intended sound for every word.

Choosing the phonemes is the first step of working with VulgarLang, my conlang development software of choice. The tusks were going to provide a "bounding box" of sorts, limiting my options for phonemes, as these prosthetics could restrict the actors' ability to speak in a normal way. Due to the limited time-scale I had, there was no opportunity to workshop sounds with the prosthetics applied, so I had to think outside the box for a solution.

At first, I explored the concept of a gestural or gesture-assisted language - a hybrid of Sign Language and spoken word that would fill in the phoneme gaps. Another member of the team had created a collection of paint-like markings, ones that could represent words, and I spent time looking at how these could be translated from movements in ink to movements of the entire arm. I felt it would be interesting to have a writing system that was functionally a scaled-down version of their gestures. This would also be an interesting challenge considering their warlike nature - could this language be "spoken" with only one hand, if the Orc is armed? Below is the "quick edition" of the first weekly presentation I gave, giving an overview of my earliest work.



At this point, however, I was finally given access to the script, and it dawned on me that this plan was, on a fundamental level, not going to work. Although Mr Brennan had very kindly offered to change some lines to make my life easier, the grammatical complexity required by the script far outstripped the capabilities of my gesture-system plan. I needed to include complex aspects of linguistics that were just not going to be possible in the work I'd already created. It was a crushing blow, but I couldn't wallow in it - it instead renewed my drive to create something truly amazing.

I went on to craft an intricate Excel spreadsheet - using CSV file tables, XLOOKUP, and a self-taught web of new formulae, I was able to create a functional "word selection" system. This allowed me to go through every Orcish line in the script, word-by-word - a much less complex task than the 4000-strong dictionaries themselves - and select which word of the four generations I preferred, with the spreadsheet automatically highlighting any duplicates. This let me see the translated lines in script order in real time - a massive boon.

With some tweaking, some swapped-out words, and a little bit of polishing, the language was ready to present to the lead actors... and they loved it! After making a dedicated document with every line translated and phonetic guides attached, I spent the last two weeks of my contract personally coaching the actors in their Orcish lines, to ensure they sounded perfect. My final day was spent recording every line for them to reference from, and I left feeling safe in the knowledge I'd done everything I could to make this Orcish language shine.

"We swore never to return. You can never trust a human. Poor workers, greedy eaters. Skinny and soft."