Hon. Pierre J. Dalphond: Dear colleagues, allow me to explain why I support Bill C-11.
I should point out that I don’t intend to comment on each clause of this bill or on the proposed amendments.
I also won’t be commenting on the important role that the Broadcasting Act has played in supporting and developing Canadian culture, whether in French, English or indigenous languages or in languages other than the official and indigenous languages. Others who spoke before me did so enthusiastically, including our new colleague, Senator Cardozo, and one of our longest-serving members, Senator Dawson.
I will only speak to one issue, which I consider to be at the heart of this bill, and that is the discoverability of Quebec and Canadian cultural products on the most well-known platforms.
First, I would like to make it clear that I do not believe that there is a vast conspiracy among platforms to make English the universal language and promote certain American values.
A 2009 UNESCO report entitled “Twelve years of measuring linguistic diversity in the Internet: balance and perspectives” noted that the presence of English on the internet had fallen from 75% in 1998 to 45% in 2005.
More recently, the Observatory of the Linguistic and Cultural Diversity in the Internet, an organization that is part of the Organisation internationale de la Francophonie, reported that in 2021, the share of English as an overall percentage of all pages available on the web was no more than 26.5%.
This shows that web content is becoming more and more diverse. The virtual warehouse, so to speak, is getting bigger and bigger and contains more and more products in different languages.
Another important statistic that stuck with me has to do with internet penetration. According to 2020 figures from internet World Stats, only 35.2% of French speakers worldwide have access to the Internet, while 77.5% of English speakers, 70.4% of Spanish speakers and 53% of Arabic speakers do.
The reason for the lack of internet access among French speakers is the low internet connection rate in French-speaking Africa, which is currently just 40%. It is estimated, however, that by 2060, internet penetration will be 85% among African French speakers. As Senator Gerba already pointed out, Africa is vital to the future of the Francophonie. There is no doubt that gradually connecting hundreds of millions of francophone Africans will create a need to produce many French-language cultural and other products, which will increase the amount of francophone content on the internet. I am delighted about that.
Of course, the existence of French-language online content is a prerequisite for the consumption of French-language cultural products. If none is available, if there is nothing on the shelves, then nothing will be consumed.
It should be noted that although English-language content no longer represents the majority of the content available — far from it — that is not the case when we look at content viewed. In fact, 61.1% of the most visited sites are in English, according to the September 2022 edition of the W3Techs Web Technology Survey.
Another study found that 85% of streams on Spotify are from 0.7% of the catalogue. There are several factors that may explain the over-consumption of certain products, including cultural products in English.
One of those factors is the smaller number of French-language platforms. That is why the member countries of the Francophonie, including Canada, launched the platform TV5MONDEplus in September 2020 to showcase French-language products. This free platform is like Netflix for the Francophonie. It enhances the online presence of television shows and movies produced in French, helps to promote the international Francophonie’s creations, and increases the discoverability of French-language content on the internet. TV5MONDEplus’s French-language productions are currently available in 196 countries.
However, another similar factor seems to account for the low consumption of French-language cultural products, and that is what are known as the platform’s suggestions.
A study found that 70% to 80% of the content watched on YouTube is based on recommendations. Users visit one page, and then they are given other recommendations and end up spending a lot of time watching.
As you know, these recommendations are made based on algorithms.
No outside experts, in either Europe or North America, have access to the details of the programming parameters of these algorithms, since the platforms consider them to be trade secrets. Europe is actually preparing regulations on this issue.
In light of this situation, researchers have begun measuring the discoverability of Quebec’s French-language cultural content on the main platforms.
In a March 2021 report entitled Être ou ne pas être découvrable?, the Université du Québec à Montréal’s research lab on discoverability and the transformation of cultural industries in the e-commerce era proposes the following definition of discoverability:
The system of “discoverability” is a set of processes that structure and determine the possibility and ability of audiences to discover cultural products online, i.e. to locate this content or have it presented to them, without necessarily searching for it in a vast database of content organized by prescription- and recommendation-based systems.
This definition emphasizes the multiple complex processes and dynamics that occur between an online consumer and a platform, as well as the impact these processes have on an audience’s propensity to discover products.
What I am about to say is an oversimplification, but basically, this complex and dynamic process is somewhat akin to product placement at your local supermarket. Often, the product that sells the best gets the spotlight. That product is strategically positioned, so there may be four or five competing products in close proximity to the featured product, but they are placed on the top or bottom shelf, where most consumers are unlikely to see them.
Sure, consumers have free choice because they’re the ones choosing the most visible, ideally positioned product, but we all know that positioning is the supermarket’s decision, whether it’s because the product’s profit margin is bigger or because the supplier paid for advantageous in-store product placement.
If the government chooses to intervene and require equitable positioning of all products, no one can seriously suggest that would impinge on consumers’ freedom of choice. One could actually argue that it gives them greater choice.
For products in an online platform’s warehouse, the shelves become algorithms. Without algorithms, these platforms would be more like massive libraries with no filing system.
These algorithms are growing increasingly sophisticated thanks to artificial intelligence. They can recognize each consumer, remember everything they’ve viewed for the past few weeks, months or even years, know how much they’re willing to pay when they make a purchase, and more. The algorithm is designed to anticipate the consumer’s latest needs and present content for their consideration.
In some cases, this is an entirely neutral operation that produces the result the consumer wants, even if the algorithm’s parameters aren’t verifiable.
As a result, according to them, interfering with the algorithm or even obtaining the details of these parameters is tantamount to threatening freedom of choice. That is what I have heard in several speeches over the past few days.
That assumes that the algorithm and its artificial intelligence are completely neutral and are capable of anticipating users’ needs in an impartial manner. That assumes that there is no possible cultural bias in the algorithm’s very complex programming.
That also assumes, of course, that there is no programming designed to boost clicks, watch time or the associated revenues.
Unfortunately, now and then, various investigations and revelations, particularly before the U.S. Congress, have proven these assumptions to be false. That is why the European Union and many other countries have decided to regulate the products offered by platforms in order to protect their country’s cultural specificity.
The scientific report published on March 8, 2021, by the research lab on discoverability that I spoke about a few moments ago shows that Canada needs to do the same for French-language cultural products. That report found that there are barriers to discoverability and pointed out the following problems.
First, there is no “Quebec category” on Netflix, iTunes, YouTube and the like.
Second, the presence of Quebec audiovisual content is very low, which explains why the algorithms do not find any or offer any. For example, none of the 29 Quebec films produced in 2016 are on Netflix. As for the 29 new films that were made during the study, 10 are on iTunes and 19 are on YouTube’s for-pay platform.
Third, there were hardly any Quebec films and shows available to stream on transnational platforms.
Fourth, new platforms such as Disney+, Amazon Prime and Apple TV contain little or no Quebec audiovisual content.
Fifth, lists of the latest Quebec songs can be found on most platforms, but they are not very visible and are rarely recommended. The situation is even worse for older hits.
Sixth, when it comes to streaming music, tests done from March to August 2019 showed that none of the “premium” streaming services met the very specific expectations of the example listener used for testing purposes.
Seventh, the platforms do not provide any details on the content consumed in Quebec or the consumption of Quebec content.
In conclusion, both the content of the platforms and the reference algorithms led to poor results for Quebec cultural products.
In this context, it only makes sense to regulate Quebec and Canadian content on these platforms, as well as the discoverability of the Quebec and Canadian products stored on them. What I said about French-language cultural tools also applies to Indigenous- and English-language Canadian cultural tools.
That’s why I support Bill C-11, which will promote Canadian content and ensure that algorithms will also present these products and make them discoverable.
Thank you for your attention. Meegwetch.