J.L.Lemke On-line Office


The NASA Earth Observatory Web

I will pass on now to a series of three webpages from a different website within the large NASA domain, the Earth Observatory site. These pages are intended for a less technically expert or scientifically knowledgeable public, as we can see by comparing its much reduced use of technical language and complex visual displays with the pages we have just analyzed. 




The Earth Observatory is a website with many, presumably deliberate, parallels to the GSFC site. Where the GSFC site is intended to allow technically expert users to access large scientific datasets of satellite images (clicking on one of the seven textlinks takes you to a page where you input the parameters of the data you want and then download large amounts of coded data), the Earth Observatory site provides a similar but greatly simplified service for students, teachers, and the general public. The page displayed here is the dataset portal page, very similar in content and design to the GSFC page. 

This time there is a set of instructions on how to use the mouse to activate dynamic images and lists of links to various datasets. Passing the mouse cursor over each of the five globe icons and their brief titles at right displays a larger more detailed globe image, presumably false-colored with data-based information and a set of links to datasets. These are very simple images with no captions. They are also tied cohesively to the introductory note for each set of displayed links by a fairly trivial repetition of lexical items. For example, the Life on Earth menu item displays text which reads: 

Life on Earth

Parameters measuring life on Earth include (follow the links to view the datasets):

 They are cohesively tied to the displayed links by thematic collocation (‘biosphere’, synonymy; ‘vegetation’, meronymy; ‘chlorophyll’, collocation), in the same way, verbally, that the displayed introduction and links are cohesive with each other. The menu icon and the displayed globe image are cohesive visually only by the similarity of shape (round) and dominant color (green). But there is a cross-modal cohesive link between the word “earth” in the menu and image of a globe of the earth, which one might call a cross-modal collocation, since this association is a conventional one produced by their frequent co-occurrence in other contexts. 

Structurally, we have a similar but simplied structure to that of the GSFC page. There are two columns with a horizontal connection of theme: topics (hyponymy or subcategorization) and a vertical list-structure with the same sort of co-hyponymy based on the atmosphere-hydrosphere-lithosphere-biosphere thematic formation used for the GSFC page. The displayed introductions and links have a similar internal structure, with a title, a specification, and the list of links. In addition there is the co-displayed globe, organizationally suggesting a thematic tie to the items in the list as well as the title. With this prompt, we can interpret the Life on Earth displayed globe as showing vegetation, green with chlorophyll, covering the landforms of Africa in an uneven way. We now read regions of the globe image as jungles and deserts, in part because the dynamic visual and interactive organization of the page implies a thematic relation to the verbal items in the links list. 

Orientationally, this page addresses the user verbally through imperatives (Proposal: demand: action). The default lefthand display gives instructions: “Click to view …” “Pass your cursor …” and the one for Life on Earth, like the others says “Follow the links …”. This constructs a Guide: Visitor relationship, with the author/source as more knowledgeable, not just about the earth or NASA data, but also about the use of the webpage itself. Otherwise we also have a [Proposal: offer: information] orientation to the user, as with the GSFC page. In the original version shown here, next to the telescope (a cross-modal collocational symbol of “observatory”) we read, “View data holdings …” which we can take to be both an offer and a demand in the sense that one can read it semantically as “if you want to view the data holdings … click this link”. The telescope image has cohesive ties to ‘view’ and ‘observatory’ in the visual paired-structure that links it to the text next to it, but no other semantic relationship. 

In terms of stance to content, there is very little content as such, so most of the evaluative dimensions are unmarked. The images show a fairly realistic, but not high resolution image for the telescope (high warrantability, but no importance) and a globe (for Life on Earth) that has the same warrantability issues discussed above for the Monsoon Dynamics image. 

I am more interested at this point in passing from the analysis of multimodality, which I think is sufficiently illustrated already, to hypertextuality. In a sense we have already seen in these dynamic pages a measure of hypertextuality within the page frame, insofar as the mouse-over display of images and text represents an interactive potential for multi-sequential display to the user of lexias or meaning units. The user can move the mouse cursor to show any of the displayed items in any order. The meanings along such traversals, beyond those of individual displays, are mainly cohesive: each element is an instance of some general category, and therefore with some thematic and/or visual similarities to the others, and as we catenate them we are cumulating toward an exhaustive exploration of the category. This is a fairly common traversal principle in websites. We are presented with a general theme or topic and links to subcategorized elements of it. We can examine these in any order, but as we examine more and more of them, we construct a traversal which is more than the sum of its parts primarily insofar as the whole also exhausts the subcategorization structure and we find we have all the parts or pieces. 

If, however, we now accept thepage’s offer of information and accede to the demand to click on a link, such as Vegetation under Life on Earth, we can consider the meaning relations between this portal page and what appears next: 


(since updated as http://earthobservatory.nasa.gov/Observatory/Datasets/ndvi.fasir.html )



In terms of Organizational meaning relations, we must first recognize that the act of clicking and following the link, i.e. calling up the display of the page is an actional-structural tie. This is similar to the cross-textual tie we find when we turn to an indicated page from one part of a book to another. Semantically, this is a two-part relation, in which the first part (anchor) in this case specifies the kind of information (dataset) to be found in the second part (target). We can consider this akin to the verbal semantic relationship of Expansion: Elaboration (between clauses). There are also of course cohesive ties, as we would expect. Verbally there is ‘vegetation’ and ‘dataset’ repeated on the new page. Visually, there is another globe, and a map of the earth, both with green coloring distributed unevenly. 

This new page has a very different compositional structure, though it can still be read as a vertical stack of information units: website banner and menu, page title, sample visual data displays, input form to specify a dataset, visual-numerical color key, “product description”. This is an invariant pattern for all the dataset access pages following from the links on the original portal page. 

The page constructs a complex source: user relationship. In addition to its implied offers of further information and demands for action (click to follow link), it also now presents a [Proposal: demand: information] in the guise of the input forms. This is the interactive equivalent of a verbal question. In order to respond to this demand, we must also accede to procedural demands for action (e.g. use a drop-down list, click on a button to submit the data form). But we are now given a large number of choices, in effect a combinatorial freedom, which is typical of the open web hypertext medium. 

This type of author: user relationship in hypertext gives rise to a sense of interactive dialogue. The page asks us to fill in the form, but by filling it in and clicking one of the “Build” buttons, it is we who are now demanding information and action, and in general the programs and databases behind the website will accede to these demands. One could say that this page offers the opportunity to make a demand. 

The “instructional” links “How to build …” are also interestingly complex in terms of the semantics of exchange relations. On the surface they offer information “here is how to build …”, but they demand an action (click the link) to get that information. When the user clicks, s/he is not just acceding to the demand for action, but also demanding the information, i.e. asking the question “How do I build ….” Or demanding “Tell me how to build …”. 


Within this page there are also some salient cross-modal ties, both structural and cohesive, between the verbal text, the visual images, and the interactive form (which is also both verbal and visual, but functionally is a hybrid medium in its own right with different semiotic and actional affordances from either pure written text or depictive images). The color key is cohesively tied to the sample data display images higher up the page. It provides an Expansion: enhancement, providing the conditions under which the image corresponds to the numerical sattelite data. The product description is an Expansion: elaboration of the color key, in part, and it also has cohesive cross-modal ties to the sample data displays, inviting us to connect visual elements to ‘greenness … soil, snow, ice …forest … crops’. 

With proper attention to the organizational conventions linking labels and titles to images and sections of text, we also find interesting cohesive links among the sample displays, the input forms, and the product description through terms for months and seasons: August 1994, start month, year, summer, winter which form a classic “similarity chain” even though this is not connected text. The chain serves to help us construct meaning relationships among the different sections of the page. If we fill in the data for building a “strip” animation, and click Build, the result is a new page, not permanently on the webserver but created just for us (figure 5). Below the usual website header and site title, there is a cohesive page title or image label “Vegetation/NDVI”, continuing the identity and similarity cohesion chains from the original portal page (vegetation) and the product description and color key image elements of the previous (anchor) page.  Below this we see a Quicktime strip image in which by dragging with the mouse along the slider bar at the bottom, we shift time backwards and forwards, with time labels beneath the image changing to index this. The rightmost and leftmost positions correspond to the input-form data we supplied for start month, end month, and year(s). What we see is the same world map, but the colorations change with the seasons. Below the animation strip-map is text which is instructional in nature; i.e. it is about the procedures for viewing the animated map, not about the thematic content of vegetation that changes with the seasons.

And below this is the color key reproduced as on the previous page, and the product description repeated for reference (so you don’t have to go back and forth between pages; it is a slow process for the server to generate the strip map and for it to be downloaded and displayed by the browser client). The instructional text sets up the same sort of source: user relationship. The map image itself has the same sort of realism and resolution levels as for the sample displays on the previous page. The strip map is to be read in relation to the color key and product description as before.

What shall we say about the cross-link meaning relation of these two pages? It is again a kind of exchange relationship (orientationally): this is the server’s answer to our question, its response to our demand, which was itself an acceptance of an offer made on the prior page. The dialogue genre is extended one turn.

Organizationally, the cross-link relationship is a more complex actional-structural one than that between the portal page and the input-form page. The action is not simply clicking the Build button, but also supplying the information in the form. There will not be a strip page (but rather an error page, or no response) otherwise. So the interactive first-part of the structural pair involves supplying information as well as demanding it. The information supplied through the input-form is verbal-numerical, but its internal organizational semantics is structured by a visual table, i.e. by a layout of relations among elements similar to that of a table in a print medium. (See Lemke 1998b for a discussion of the historical and semantic relations of tables to connected text and to more graphical visual displays.)

Presentationally and thematically, the inter-page relationship would seem to be that of exemplification. The first-part page offers a template for a category of possible strip-maps, and the second-part displays one instance from this category. This is the primary meaning relationship, but secondarily one can also say that the new strip map provides an Extension: additive to the information in the sample display. As if to say: In August 1994 the world looked like this, and in other months (e.g. April 1994) it looked like this.

It would also be possible, depending on the input parameters, to obtain a relation of temporal meronymy: the sample display shows one month within the whole year (or period) which the strip-map shows, thus including it as a part.

In purely visual terms, we could compare the two strip maps (sample and animation), to see a cohesive tie (the same continents and landforms) and a spatial-distributional difference in colors. There would be a very different pattern of salience and eye-movement across the two maps. How could we name their visual meaning-relation? It is somewhat akin to “morphing”, a constant general form with varying details (or the continuous passage from one image to another). It is also somewhat like the relation of theme-and-variations in music: a constant overall form, with varying lesser details. This assumes that we see, culturally, configurations of shapes and masses as more significant visually than relatively minor changes within a still relatively constant color palette. This visual relationship of the pages seems subordinated to the semantic relationships already discussed.

There is of course another visual relationship of the pages: their sequential display in the same visual space, like the frames of an animation or film. This is a dynamic visual relationship, in time more than in space, in the window of the browser.

Some hypertext systems, and even the web, can launch new windows for successor pages, and display these side by side, or in other spatial relationships. So there is a potential significance in the choice to replace one page by another in the same external visual space. This potential does not seem to be exploited here in any marked way.