By Dr. Ryan Rumpf

In the last article, we explored what some qualities of effective language instruction looks like in practice. I ended the article by saying that there is no actual mention of any of those strategies or concepts in our lovely compliance mandates. Federal law, largely based on major court decisions, requires that schools provide an “educationally sound” language instruction program, staff it and provide sufficient resources, and periodically determine if it is “working.” It’s a pretty broad stroke.

Clearly, it leaves a lot open to interpretation. Over  the years, I have seen and heard some very strong feelings over what a language instruction education program should look like. I’m certain this article may push a few buttons. There have been times in my career when I wholeheartedly believed in one specific way to do ELD too. Whether I was doing a service delivery model, something more collaborative, or a self-contained program, after time, something just started feeling wrong or I found a school where it simply wouldn’t work. This article will explore the first part of the mandate, educationally sound programming, while acknowledging that any program will fail if there is not sufficient staff, resources, training and of course, effective leadership. 

Think back to the core qualities of effective language instruction: 1) the students understand the language in front of them (input), 2) they have opportunities to use language (output) and 3) they are presented with and free to use all their linguistic resources (multilingualism). Now consider the school day. Depending on where you are, the school day is somewhere between 6-8 hours. Considering breaks, passing periods, lunch, etc., how many of those hours are dedicated to academic instruction? And during how many of those hours should students have access to those qualities of effective language instruction? I think it is reasonable to not only expect the entire time, but I personally feel that it should be the foundation of what is considered “educationally sound.” 

I know some say it’s not possible in their situation. But let’s take a moment to step back and look at the program models. As I briefly touched upon in the first article, bilingual education has been around for centuries. Today, bilingual education can come in many forms. There are transitional programs that gradually shift students from instruction in their home language toward predominantly English instruction over time. I honestly can’t think of many better examples of “taking kids from where they are to where they need to be.” 

Then there is the incredibly popular 50/50 two-way, dual language immersion programs filled with native speakers of both languages. In these programs, half of the day is in one language and half the day is in the other. Parents love these – especially English-speaking parents. In fact, in some places, their popularity among English-speaking families has created a different problem: demand is so high that multilingual learners and heritage speakers may struggle to get in. 

Analyze these models with our list of qualities of effective language programs and they have the potential to check all of the boxes. Importantly, these programs check those boxes throughout the school day, so I consider them “educationally sound.” There have been major longitudinal studies that show students who were in dual language or bilingual programs outperform students that were in English-only programs over time. But for some reason, most districts across the country are reluctant to even try. The reason? They can’t find the staff, there are too many languages in the district, or there are not enough students. But here’s the thing you won’t find in textbooks: the entire school does not have to operate this model. Simply as an example, and there are others, let’s say a school has three to four cohorts (classes) of students per grade level – one of them can be the dual language classroom. Personally, I think they are more possible than we’d like to admit, but there are some English-only models that check the boxes too. 

Back in the days after the courts and legislature got involved in all of this, districts had to pick a model. For the schools that did not go the bilingual route, pull-out ESL classes were the go-to. After some time, researchers, leaders and advocates noticed that those “services” did not do much to help the graduation or college-going rates. The problem wasn’t with providing an English class. I personally love language learning classes. The problem was with the rest of the day. If a student is pulled from a class, they are missing the opportunity to learn that material. The other problem was that an ESL does not magically help students understand what is happening in math, science, history, language arts, etc. When the students were off on their own, they were in a “sink or swim” situation. This method checked our boxes for one class out of the day. That’s a problem and it’s one that persists across most school districts today. 

This doesn’t mean pull-out ELD is inherently bad. There are times when students benefit enormously from protected time focused specifically on English development. The problem occurs when that class is the program and it takes the place of other important learning opportunities. Forty-five minutes of excellent language instruction cannot make the other five or six hours of inaccessible instruction “educationally sound.” And kids should not be forced to miss math or science for an English lesson. ELD can and should be an important component of a language instruction program, but it’s not the entire program. 

So what are other English-only models. Drawing from the research I mentioned in the last article, models referred to as “sheltered instruction” started to become common. The idea of using techniques such as pictures, demonstrations, adjusted speech, etc., to make academic content “comprehensible” was more appealing than pulling students from those classes to teach English grammar, literacy, etc. In the late 1980s and 1990s, schools in California started experimenting with a model called Specially Designed Academic Instruction in English (SDAIE). The goal was to use specific techniques, such as contextualization, modeling, and scaffolding to make grade level content comprehensible to multilingual learners. In the original sheltered methods and SDAIE, sources advised that students acquire an intermediate level of English proficiency to get the most out of these classes. Either way, they were better than the alternative. 

Then a group of researchers from California developed the Sheltered Instruction Observation Protocol (SIOP) which synthesized all of the varying methods and ideas into one cohesive model. It eventually emerged as the go-to model for sheltered instruction, despite Krashen’s warnings. There are 8 components and 30 features that bring the whole thing together. They claim that it can be beneficial for multilingual learners at any proficiency level. It checks all of our boxes for effective language acquisition and when implemented well, can provide students with a meaningful experience all school day. It can be “educationally sound” if, and this is a big if, it is done across content areas. Another challenge is implementation. It’s hard enough to get SIOP onto the PD schedule, let alone convince any teacher that sees an EL to use 8 components and 30 features consistently, especially when there are other mandates, other students and a limited amount of time. The other issue is the fluid and ever-changing nature of the EL population. 

Students acquiring English go through a series of predictable stages on their English journeys. At each stage, they need to stretch their abilities so they keep moving down their English path. So, a student who is three years into their journey needs different kinds of experiences and interactions than a student that is just taking their first steps. Imagine an 8th grader who just arrived in the US and will be speaking their first words in English on their first day of school (maybe). You can have a beautifully implemented SIOP model for all core content areas. You can provide visuals, gestures, modeling, language objectives, sentence frames, partner interaction, adapted text, background building, supplementary materials, and comprehensible teacher speech. Much of that day still has to push the larger portion of class who are further along in their English journey, which will leave our new 8th grader wondering what is going on. 

This isn’t a criticism of SIOP or the classroom teacher. In fact, I would want every teacher who works with newcomers to know and use these strategies. The problem is structural. It happens in the best bilingual programs as well. A student entering a long-running dual-language program in eighth grade may not have the proficiency in one or both languages that the curriculum assumes. A teacher cannot provide grade-level instruction to an entire class and simultaneously calibrate every interaction to a student who is taking their first steps in English. At some point, we have to stop asking individual teachers to solve a program-design problem. It can and will burn people out on these valuable ideas. Frankly, in some places, it already has. 

Enter newcomer programs. A strong newcomer program concentrates language development, comprehensible grade-level content, literacy when needed, orientation to a new school system, targeted counseling, social work and importantly, opportunities to build relationships and belonging. The point isn’t to put students somewhere until they “learn enough English.” Students are still learning math, science, social studies and language arts; we’re simply designing more of their day around their current linguistic starting point. These programs, whether implemented as a “center” or a school-within-a-school are able to focus on all of these things for a limited time before sending students on their way after they have reached a defined set of characteristics. Notice I didn’t put a time marker on there; it’s about where students are in their English journey. 

As I mentioned before, every district is different. And just like bilingual programs, newcomer programs don’t have to be a big glorious building either. Imagine something like: 

Low-Incidence Newcomer Programs: Maybe there are only a few newcomers. Perhaps those students spend one or two periods together receiving core instruction, literacy, orientation support and spend the rest of their day in mainstream classrooms where the teacher has been SIOP trained.

Moderate-Incidence Newcomer Programs: Create a newcomer strand or cohort. Students should share the same ELD classes and have dedicated sheltered content courses built into the master schedule. Students still take electives with their peers; perhaps, one elective is focused on cultural exchanges that familiarize students with their new community. 

High-Incidence Newcomer Centers: These are legitimate options. You can concentrate your specialized content and ELD teachers as opposed to hiring more of them and spreading them out. You can hire counselors, family support, specialized social workers, etc. that can comprehensively meet students needs in one spot. Asking each individual building to offer these things spreads your specialized and valuable staff thin, increases burnout and falls far short of what a centralized place can do. 

Of course, newcomer centers have legitimate risks. A newcomer center or program cannot become a place where students stay indefinitely. At that point, a valuable program risks becoming segregation. Newcomer centers are temporary, purpose-driven, and a major boost to students just beginning a new life, in a new school, in a new language. The goal is to provide a soft landing for students as they arrive, build them up, then send them on their merry way. 

There are certainly other models, and as I mentioned, I have tried several on my journey. I have pulled students out. I have pushed in. I was on the co-teaching train for a decade and even trained people to do it. These models can work incredibly well. I’m less concerned with the label than whether the model checks all three boxes throughout the school day. Any student in any classroom should be able to 1) understand what they are learning (input), 2) have opportunities to talk through and make sense of the content (output) and 3) use their full linguistic repertoires (multilingualism). It should happen wherever students are. Anything less is something to improve. 

So what would I do if I got to start something from scratch? Luckily, I was able to do that once and that’s where we’re going next. 

To be continued…